MOSEK

The MOSEK optimization tools
manual.
Version 6.0 (Revision 61).
ii
Published by MOSEK ApS, Denmark.
Copyright (c) 1998-2010 MOSEK ApS, Denmark. All rights reserved..
Disclaimer: MOSEK ApS (the author of MOSEK) accepts no responsibility for damages resulting
from the use of the MOSEK software and makes no warranty, neither expressed nor implied,
including, but not limited to, any implied warranty of tness for a particular purpose. The
software is provided as it is, and you, its user, assume all risks when using it.
Contact information
Phone +45 3917 9907
Fax +45 3917 9823
WEB http://www.mosek.com
Email sales@mosek.com Sales, pricing, and licensing.
support@mosek.com Technical support, questions and bug reports.
info@mosek.com Everything else.
Mail MOSEK ApS
C/O Symbion Science Park
Fruebjergvej 3, Box 16
2100 Copenhagen
Denmark
iii
iv
Contents
1 Changes and new features in MOSEK 3
1.1 Compilers used to build MOSEK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 General changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Optimizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Interior point optimizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.2 The simplex optimizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.3 Mixed-integer optimizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 License system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 Other changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.6 Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.7 Platform changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 The MOSEK optimization tools 7
2.1 What is MOSEK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.1 Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 How to use this manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3 Getting support and help 9
3.1 MOSEK documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 Additional reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4 Using the MOSEK command line tool 11
4.1 Getting started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.2.1 Linear optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.2.2 Quadratic optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2.3 Conic optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.3 Passing options to the command line tool . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.4 Reading and writing problem data les . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.4.1 Reading compressed data les . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.4.2 Converting from one format and to another . . . . . . . . . . . . . . . . . . . . . 18
4.5 Hot-start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.5.1 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.6 Further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.7 Solution le ltering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
v
vi CONTENTS
5 MOSEK and AMPL 21
5.1 Invoking the AMPL shell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.2 Applicability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.4 Determining the outcome of an optimization . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.5 Optimizer options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.5.1 The MOSEK parameter database . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.5.2 Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.6 Constraint and variable names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.7 Which solution is returned to AMPL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.8 Hot-start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.9 Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.10 Using the command line version of the AMPL interface . . . . . . . . . . . . . . . . . . 28
6 MOSEK and GAMS 29
7 MOSEK and MATLAB 31
8 Interfaces to MOSEK 33
8.1 The optimizer API . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
9 Modelling 35
9.1 Linear optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
9.1.1 Duality for linear optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
9.1.2 Primal and dual infeasible case . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
9.2 Quadratic and quadratically constrained optimization . . . . . . . . . . . . . . . . . . . 38
9.2.1 A general recommendation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
9.2.2 Reformulating as a separable quadratic problem . . . . . . . . . . . . . . . . . . 39
9.3 Conic optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
9.3.1 Duality for conic optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
9.3.2 Infeasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
9.3.3 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
9.3.4 Potential pitfalls in conic optimization . . . . . . . . . . . . . . . . . . . . . . . . 46
9.4 Nonlinear convex optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
9.4.1 Duality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
9.5 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
9.5.1 Avoid near infeasible models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
9.6 Examples continued . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
9.6.1 The absolute value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
9.6.2 The Markowitz portfolio model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
10 The optimizers for continuous problems 55
10.1 How an optimizer works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
10.1.1 Presolve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
10.1.2 Dualizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
10.1.3 Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
10.1.4 Using multiple CPUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
CONTENTS vii
10.2 Linear optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
10.2.1 Optimizer selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
10.2.2 The interior-point optimizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
10.2.3 The simplex based optimizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
10.2.4 The interior-point or the simplex optimizer? . . . . . . . . . . . . . . . . . . . . . 63
10.2.5 The primal or the dual simplex variant? . . . . . . . . . . . . . . . . . . . . . . . 64
10.3 Linear network optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
10.3.1 Network ow problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
10.3.2 Embedded network problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
10.4 Conic optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
10.5 Nonlinear convex optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
10.6 Solving problems in parallel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
10.6.1 Thread safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
10.6.2 The parallelized interior-point optimizer . . . . . . . . . . . . . . . . . . . . . . . 67
10.6.3 The concurrent optimizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
10.7 Understanding solution quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
10.7.1 The solution summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
11 The optimizer for mixed integer problems 73
11.1 Some notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
11.2 An important fact about integer optimization problems . . . . . . . . . . . . . . . . . . 74
11.3 How the integer optimizer works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
11.3.1 Presolve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
11.3.2 Heuristic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
11.3.3 The optimization phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
11.4 Termination criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
11.5 How to speed up the solution process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
11.6 Understanding solution quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
11.6.1 Solutionsummary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
12 The analyzers 79
12.1 The problem analyzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
12.1.1 General characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
12.1.2 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
12.1.3 Linear constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
12.1.4 Constraint and variable bounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
12.1.5 Quadratic constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
12.1.6 Conic constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
12.2 Analyzing infeasible problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
12.2.1 Example: Primal infeasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
12.2.2 Locating the cause of primal infeasibility . . . . . . . . . . . . . . . . . . . . . . 84
12.2.3 Locating the cause of dual infeasibility . . . . . . . . . . . . . . . . . . . . . . . . 85
12.2.4 The infeasibility report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
12.2.5 Theory concerning infeasible problems . . . . . . . . . . . . . . . . . . . . . . . . 89
viii CONTENTS
12.2.6 The certicate of primal infeasibility . . . . . . . . . . . . . . . . . . . . . . . . . 89
12.2.7 The certicate of dual infeasibility . . . . . . . . . . . . . . . . . . . . . . . . . . 90
13 Feasibility repair 91
13.1 The main idea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
13.2 Feasibility repair in MOSEK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
13.2.1 Usage of negative weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
13.2.2 Automatical naming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
13.2.3 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
14 Sensitivity analysis 95
14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
14.2 Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
14.3 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
14.4 Sensitivity analysis for linear problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
14.4.1 The optimal objective value function . . . . . . . . . . . . . . . . . . . . . . . . . 95
14.4.2 The basis type sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 97
14.4.3 The optimal partition type sensitivity analysis . . . . . . . . . . . . . . . . . . . 97
14.4.4 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
14.5 Sensitivity analysis with the command line tool . . . . . . . . . . . . . . . . . . . . . . . 102
14.5.1 Sensitivity analysis specication le . . . . . . . . . . . . . . . . . . . . . . . . . 102
14.5.2 Example: Sensitivity analysis from command line . . . . . . . . . . . . . . . . . . 103
14.5.3 Controlling log output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
A MOSEK command line tool reference 105
A.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
A.2 Command line arguments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
A.3 The parameter le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
A.3.1 Using the parameter le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
B The MPS le format 109
B.1 The MPS le format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
B.1.1 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
B.1.2 NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
B.1.3 OBJSENSE (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
B.1.4 OBJNAME (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
B.1.5 ROWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
B.1.6 COLUMNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
B.1.7 RHS (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
B.1.8 RANGES (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
B.1.9 QSECTION (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
B.1.10 BOUNDS (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
B.1.11 CSECTION (optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
B.1.12 ENDATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
B.2 Integer variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
B.3 General limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
B.4 Interpretation of the MPS format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
CONTENTS ix
B.5 The free MPS format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
C The LP le format 121
C.1 A warning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
C.2 The LP le format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
C.2.1 The sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
C.2.2 LP format peculiarities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
C.2.3 The strict LP format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
C.2.4 Formatting of an LP le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
D The OPF format 129
D.1 Intended use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
D.2 The le format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
D.2.1 Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
D.2.2 Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
D.2.3 Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
D.3 Parameters section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
D.4 Writing OPF les from MOSEK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
D.5 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
D.5.1 Linear example lo1.opf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
D.5.2 Quadratic example qo1.opf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
D.5.3 Conic quadratic example cqo1.opf . . . . . . . . . . . . . . . . . . . . . . . . . . 137
D.5.4 Mixed integer example milo1.opf . . . . . . . . . . . . . . . . . . . . . . . . . . 138
E The XML (OSiL) format 141
F The solution le format 143
F.1 The basic and interior solution les . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
F.2 The integer solution le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
G The ORD le format 145
G.1 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
H Parameters reference 147
H.1 Parameter groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
H.1.1 Logging parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
H.1.2 Basis identication parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
H.1.3 The Interior-point method parameters. . . . . . . . . . . . . . . . . . . . . . . . . 149
H.1.4 Simplex optimizer parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
H.1.5 Primal simplex optimizer parameters. . . . . . . . . . . . . . . . . . . . . . . . . 153
H.1.6 Dual simplex optimizer parameters. . . . . . . . . . . . . . . . . . . . . . . . . . 154
H.1.7 Network simplex optimizer parameters. . . . . . . . . . . . . . . . . . . . . . . . 154
H.1.8 Nonlinear convex method parameters. . . . . . . . . . . . . . . . . . . . . . . . . 154
H.1.9 The conic interior-point method parameters. . . . . . . . . . . . . . . . . . . . . 155
H.1.10 The mixed-integer optimization parameters. . . . . . . . . . . . . . . . . . . . . . 155
H.1.11 Presolve parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
H.1.12 Termination criterion parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . 158
x CONTENTS
H.1.13 Progress call-back parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
H.1.14 Non-convex solver parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
H.1.15 Feasibility repair parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
H.1.16 Optimization system parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
H.1.17 Output information parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
H.1.18 Extra information about the optimization problem. . . . . . . . . . . . . . . . . . 163
H.1.19 Overall solver parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
H.1.20 Behavior of the optimization task. . . . . . . . . . . . . . . . . . . . . . . . . . . 165
H.1.21 Data input/output parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
H.1.22 Analysis parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
H.1.23 Solution input/output parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . 171
H.1.24 Infeasibility report parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
H.1.25 License manager parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
H.1.26 Data check parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
H.1.27 Debugging parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
H.2 Double parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
H.3 Integer parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
H.4 String parameter types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
I Symbolic constants reference 281
I.1 Constraint or variable access modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
I.2 Function opcode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
I.3 Function operand type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
I.4 Basis identication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
I.5 Bound keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
I.6 Species the branching direction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
I.7 Progress call-back codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
I.8 Types of convexity checks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
I.9 Compression types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
I.10 Cone types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
I.11 CPU type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
I.12 Data format types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
I.13 Double information items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
I.14 Double parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
I.15 Feasibility repair types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
I.16 License feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
I.17 Integer information items. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
I.18 Information item types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
I.19 Input/output modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
I.20 Integer parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
I.21 Language selection constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
I.22 Long integer information items. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
I.23 Mark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
I.24 Continuous mixed-integer solution type . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
I.25 Integer restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
I.26 Mixed-integer node selection types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
CONTENTS xi
I.27 MPS le format type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
I.28 Message keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
I.29 Network detection method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
I.30 Objective sense types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
I.31 On/o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
I.32 Optimizer types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
I.33 Ordering strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
I.34 Parameter type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
I.35 Presolve method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
I.36 Problem data items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
I.37 Problem types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
I.38 Problem status keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
I.39 Interpretation of quadratic terms in MPS les . . . . . . . . . . . . . . . . . . . . . . . . 333
I.40 Response codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
I.41 Response code type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
I.42 Scaling type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
I.43 Scaling type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
I.44 Sensitivity types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
I.45 Degeneracy strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
I.46 Exploit duplicate columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
I.47 Hot-start type employed by the simplex optimizer . . . . . . . . . . . . . . . . . . . . . 354
I.48 Problem reformulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
I.49 Simplex selection strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
I.50 Solution items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
I.51 Solution status keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
I.52 Solution types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
I.53 Solve primal or dual form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
I.54 String parameter types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
I.55 Status keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
I.56 Starting point types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
I.57 Stream types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
I.58 Integer values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
I.59 Variable types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
I.60 XML writer output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
J Problem analyzer examples 363
J.1 air04 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
J.2 arki001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
J.3 Problem with both linear and quadratic constraints . . . . . . . . . . . . . . . . . . . . . 365
J.4 Problem with both linear and conic constraints . . . . . . . . . . . . . . . . . . . . . . . 367
xii CONTENTS
License agreement
Before using the MOSEK software, please read the license agreement available in the distribution at
mosek\6\license.pdf
1
2 CONTENTS
Chapter 1
Changes and new features in
MOSEK
The section presents improvements and new features added to MOSEK in version 6.0.
1.1 Compilers used to build MOSEK
MOSEK has been build with the compiler shown in Table 1.1.
Platform C compiler
linux32x86 Intel C 11.0 (gcc 4.3, glibc 2.3.4)
linux64x86 Intel C 11.0 (gcc 4.3, glibc 2.3.4)
osx32x86 Intel C 11.1 (gcc 4.0)
osx64x86 Intel C 11.1 (gcc 4.0)
solaris32x86 Sun Studio 12
solaris64x86 Sun Studio 12
win32x86 Intel C 11.0 (VS 2005)
win64x86 Intel C 11.0 (VS 2005)
Table 1.1: Compiler version used to build MOSEK
.
1.2 General changes
A problem analyzer is now available. It generates an simple report with of statisics and informa-
tion about the optimization problem and relevant warnings about the problem formulation are
included.
A solution analyzer is now available.
3
4 CHAPTER 1. CHANGES AND NEW FEATURES IN MOSEK
All timing measures are now wall clock times
MOSEK employs version 1.2.3 of the zlib library.
MOSEK employs version 11.6.1 of the FLEXnet licensing tools.
The convexity of quadratic and quadratic constrained optimization is checked explicitly.
On Windows all DLLs and EXEs are now signed.
On all platforms the Jar les are signed.
MOSEK no longer deals with ctrl-c. The user is responsible for terminating MOSEK in the
callback.
1.3 Optimizers
1.3.1 Interior point optimizer
The speed and stability of interior-point optimizer for linear problems has been improved.
The speed and stability of the interior-point optimizer for conic problems has been improved. In
particular, it is much better at dealing with primal or dual infeasible problems.
1.3.2 The simplex optimizers
Presolve is now much more eective for simplex optimizers hot-starts.
1.3.3 Mixed-integer optimizer
The stopping criteria for the mixed-integer optimizer have been changed to conform better with
industry standards.
1.4 License system
The license conditions have been relaxed, so that a license is shared among all tasks using a
single environment. This means that running several optimizations in parallel will only consume
one license, as long as the associated tasks share a single MOSEK environment. Please note this
is NOT useful when using the MATLAB parallel toolbox.
By default a license remains checked out for the lifetime of the environment. This behavior can
be changed using the parameter MSK IPAR CACHE LICENSE.
Flexlm has been upgraded to version 11.6 from version 11.4.
1.5 Other changes
The documentation has been improved.
1.6. INTERFACES 5
1.6 Interfaces
The AMPL interface has been augmented so it is possible to pass an initial (feasible) integer
solution to mixed-integer optimizer.
The AMPL interface is now capable of reading the constraint and variable names if they are
avialable.
1.7 Platform changes
MAC OSX on the PowerPC platform is no longer supported.
Solaris on the SPARC platform is no longer supported.
MAC OSX is supported on Intel 64 bit X86 i.e. osx64x86.
Add support for MATLAB R2009b.
6 CHAPTER 1. CHANGES AND NEW FEATURES IN MOSEK
Chapter 2
The MOSEK optimization tools
2.1 What is MOSEK
MOSEK is a software package for solving mathematical optimization problems.
The core of MOSEK consists of a number of optimizers that can solve various optimization prob-
lems. The problem clases MOSEK is designed to solve are:
Linear problems.
Conic quadratic problems. (also known as second order optimization).
General convex problems. In particular, MOSEK is wellsuited for:
Convex quadratic problems.
Convex quadratically constrained problems.
Geometric problems (posynomial case).
Integer problems, i.e. problems where some of the variables are constrained to integer values.
These problem classes can be solved using an appropriate optimizer built into MOSEK:
Interior-point optimizer for all continuous problems.
Primal or dual simplex optimizer for linear problems.
Conic interior-point optimizer for conic quadratic problems.
Mixed-integer optimizer based on a branch and cut technology.
All the optimizers available in MOSEK are built for solving large-scale sparse problems and have
been extensively tuned for stability and performance.
7
8 CHAPTER 2. THE MOSEK OPTIMIZATION TOOLS
2.1.1 Interfaces
There are several ways to interface with MOSEK:
Files:
MPS format: MOSEK reads the industry standard MPS le format for specifying (mixed
integer) linear optimization problems. Moreover an MPS le can also be used to specify
quadratic, quadratically constrained, and conic optimization problems.
LP format: MOSEK can read and write the CPLEX LP format with some restrictions.
OPF format: MOSEK also has its own text based format called OPF. The format is closely
related to the LP but is much more robust in its specication
APIs: MOSEK can also invoked from various programming languages.
C/C++,
C# (plus other .NET languages),
Delphi,
Java and
Python.
Thrid party programs:
AMPL: MOSEK can easily be used from the modeling language AMPL
1
which is a high-
level modeling language that makes it possible to formulate optimization problems in a
language close to the original pen and paper model formulation.
MATLAB: When using the MOSEK optimization toolbox for Matlab the functionality of
MOSEK can easily be used within MATLAB.
2.2 How to use this manual
This manual consists of two parts each consisting of several chapters.
The rst part consists of the Chapters 4 to 14 and is a Users guide which provides a quick
introduction to the usage of MOSEK. The last part consists of appendixes A - I is a reference manual
for the MOSEK command line tool, le formats and parameters.
1
See http://www.ampl.com for further information.
Chapter 3
Getting support and help
3.1 MOSEK documentation
For an overview of the available MOSEK documentation please see
mosek\6\help\index.html
in the distribution.
3.2 Additional reading
In this manual it is assumed that the reader is familiar with mathematics and in particular math-
ematical optimization. Some introduction to linear programming is found in books such as Linear
programming by Chvatal [12] or Computer Solution of Linear Programs by Nazareth [18]. For more
theoretical aspects see e.g. Nonlinear programming: Theory and algorithms by Bazaraa, Shetty, and
Sherali [10]. Finally, the book Model building in mathematical programming by Williams [22] pro-
vides an excellent introduction to modeling issues in optimization.
Another useful resource is Mathematical Programming Glossary available at
http://glossary.computing.society.informs.org
9
10 CHAPTER 3. GETTING SUPPORT AND HELP
Chapter 4
Using the MOSEK command line
tool
This chapter introduces the MOSEK command line tool which allows the user to solve optimization
problems specied in a text le. The main reasons to use the command line tool are
to solve small problems by hand, and
as a debugging tool for large problems generated by other programs.
4.1 Getting started
The syntax for the mosek command line tool is
mosek [options] filename
[options] are some options which modify the behavior of MOSEK such as whether the optimization
problem is minimized or maximized. filename is the name of the le which contains the problem
data. E.g the
mosek -min afiro.mps
command line tells MOSEK to read data from the afiro.mps le and to minimize the objective
function.
By default the solution to the optimization problem is stored in the les afiro.sol and afiro.bas.
The .sol and .bas les contains the interior and basis solution respectively. For problems with integer
variables the solution is written to a le with the extension .int.
For a complete list of command line parameters type
mosek -h
or see Appendix A.
11
12 CHAPTER 4. USING THE MOSEK COMMAND LINE TOOL
4.2 Examples
Using several examples we will subsequently demonstrate how to use the MOSEK command line tool.
4.2.1 Linear optimization
A linear optimization problem is a problem where a linear objective function is optimized subject to
linear constraints. An example of a linear optimization problem is
minimize 10x
1
9x
2
,
subject to 7/10x
1
+ 1x
2
630,
1/2x
1
+ 5/6x
2
600,
1x
1
+ 2/3x
2
708,
1/10x
1
+ 1/4x
2
135,
x
1
, x
2
0.
(4.1)
The solution of the example (4.1) using MOSEK consists of three steps:
Creating an input le describing the problem.
Optimizing the problem using MOSEK.
Viewing the solution reports.
The input le for MOSEK is a plain text le containing a description of the problem and it must
be in either the MPS, the LP, or the OPF format. Below we present the example encoded as an OPF
le:
[comment]
Example lo1.mps converted to OPF.
[/comment]
[hints]
# Give a hint about the size of the different elements in the problem.
# These need only be estimates, but in this case they are exact.
[hint NUMVAR] 2 [/hint]
[hint NUMCON] 4 [/hint]
[hint NUMANZ] 8 [/hint]
[/hints]
[variables]
# All variables that will appear in the problem
x1 x2
[/variables]
[objective minimize obj]
- 10 x1 - 9 x2
[/objective]
4.2. EXAMPLES 13
[constraints]
[con c1] 0.7 x1 + x2 <= 630 [/con]
[con c2] 0.5 x1 + 0.8333333333 x2 <= 600 [/con]
[con c3] x1 + 0.66666667 x2 <= 708 [/con]
[con c4] 0.1 x1 + 0.25 x2 <= 135 [/con]
[/constraints]
[bounds]
# By default all variables are free. The following line will
# change this to all variables being nonnegative.
[b] 0 <= * [/b]
[/bounds]
For details on the syntax of the OPF format please consult Appendix D.
After the input le has been created, the problem can be optimized. Assuming that the input le
has been given the name lo1.opf, then the problem is optimized using the command line
mosek lo1.opf
Two solution report les lo1.sol and lo1.bas are generated where the rst le contains the interior
solution and the second le contains the basic solution. In this case the lo1.bas le has the format:
NAME : EXAMPLE
PROBLEM STATUS : PRIMAL_AND_DUAL_FEASIBLE
SOLUTION STATUS : OPTIMAL
OBJECTIVE NAME : obj
PRIMAL OBJECTIVE : -7.66799999e+003
DUAL OBJECTIVE : -7.66799999e+003
CONSTRAINTS
INDEX NAME AT ACTIVITY LOWER LIMIT UPPER LIMIT DUAL LOWER DUAL UPPER
1 c1 UL 6.30000000e+002 NONE 6.30000000e+002 0.00000000e+000 4.37499996e+000
2 c2 BS 4.80000000e+002 NONE 6.00000000e+002 0.00000000e+000 0.00000000e+000
3 c3 UL 7.08000000e+002 NONE 7.08000000e+002 0.00000000e+000 6.93750003e+000
4 c4 BS 1.17000000e+002 NONE 1.35000000e+002 0.00000000e+000 0.00000000e+000
VARIABLES
1 x1 BS 5.39999998e+002 0.00000000e+000 NONE 0.00000000e+000 0.00000000e+000
2 x2 BS 2.52000001e+002 0.00000000e+000 NONE 0.00000000e+000 0.00000000e+000
The interpretation of the solution le should be obvious. E.g the optimal values of x1 and x2 are
539.99 and 252.00 respectively. A detailed discussion of the solution le format is given in Appendix
F.
4.2.2 Quadratic optimization
An example of a quadratic optimization problem is
minimize x
2
1
+ 0.1x
2
2
+ x
2
3
x
1
x
3
x
2
subject to 1 x
1
+ x
2
+ x
3
,
x 0.
(4.2)
The problem is a quadratic optimization problem because all the constraints are linear and the objective
can be stated on the form
0.5x
T
Qx + c
T
x
where in this particular case we have that
Q =
_
_
2 0 1
0 0.2 0
1 0 2
_
_
and c =
_
_
0
1
0
_
_
. (4.3)
MOSEK assumes that Q is symmetric and positive semi-denite. If these assumptions are not
satised, MOSEK will most likely not compute a valid solution. Recall a matrix is symmetric if it
satises the condition
Q = Q
T
and it is positive semi-denite if
x
T
Qx 0, for all x.
An OPF le specifying the example can have the format:
[comment]
Example qo1.mps converted to OPF.
[/comment]
[hints]
[/hints]
[variables]
x1 x2 x3
[/variables]
# The quadratic terms are often multiplied by 1/2,
# but this is not required.
- x2 + 0.5 ( 2 x1 ^ 2 - 2 x3 * x1 + 0.2 x2 ^ 2 + 2 x3 ^ 2 )
[/objective]
[constraints]
[con c1] 1 <= x1 + x2 + x3 [/con]
[/constraints]
[bounds]
[b] 0 <= * [/b]
[/bounds]
Please note that the quadratic terms in objective are stated very naturally in the OPF format as
follows
- x2 + 0.5 ( 2 x1 ^ 2 - 2 x3 * x1 + 0.2 x2 ^ 2 + 2 x3 ^ 2 )
4.2. EXAMPLES 15
The example is solved using the
mosek qo1.opf
command line. In this case only one solution le named qo1.sol is produced. A .bas le is only
produced for linear problems.
4.2.3 Conic optimization
Conic optimization is a generalization of linear optimization which allows the formulation of nonlinear
convex optimization problems.
The main idea in conic optimization is to include constraints of the form
x
t
(
in the optimization problem where x
t
consists of a subset of the variables and ( is a convex cone.
Recall that ( is a convex cone if and only if ( is a convex set and
x ( x ( for all 0.
MOSEK cannot handle arbitrary conic constraints, only the two types
_
_
_
x R
n+1
: x
1

_
n+1
j=2
x
2
j
_
_
_
(4.4)
and _
_
_
x R
n+2
: 2x
1
x
2

n+2
j=2
x
2
j
, x
1
, x
2
0
_
_
_
. (4.5)
(4.4) is called a quadratic cone whereas (4.5) is called a rotated quadratic cone.
Consider the problem
minimize 1x
1
+ 2x
2
subject to
1
x1
+
2
x2
5,
x 0
(4.6)
which may not initially look like a conic optimization problem. It can however be reformulated to
minimize 1x
1
+ 2x
2
subject to 2x
3
+ 4x
4
= 5,
x
2
5
2x
1
x
3
,
x
2
6
2x
2
x
4
,
x
5
= 1,
x
6
= 1,
x 0.
(4.7)
Problem (4.6) and problem (4.7) are equivalent because the constraints of (4.7)
x
2
5
x
1
=
1
x
1
2x
3
and
x
2
6
x
2
1
x
2
2x
4
and hence
1
x
1
+
2
x
2
2x
3
+ 4x
4
= 5.
The problem (4.7) is a conic quadratic optimization problem.
Using the MOSEK OPF format the problem can be represented as follows:
[comment]
Example cqo1.mps converted to OPF.
[/comment]
[hints]
[/hints]
[variables]
x1 x2 x3 x4 x5 x6
[/variables]
x1 + 2 x2
[/objective]
[constraints]
[con c1] 2 x3 + 4 x4 = 5 [/con]
[/constraints]
[bounds]
# We let all variables default to the positive orthant
[b] 0 <= * [/b]
# ... and change those that differ from the default.
[b] x5,x6 = 1 [/b]
# We define two rotated quadratic cones
# k1: 2 x1 * x3 >= x5^2
[cone rquad k1] x1, x3, x5 [/cone]
# k2: 2 x2 * x4 >= x6^2
[cone rquad k2] x2, x4, x6 [/cone]
[/bounds]
For details on the OPF format please consult Appendix D. Finally, the example can be solved using
the
mosek cqo1.opf
command line call and the solution can be studied by inspecting the cqo1.sol le.
4.3. PASSING OPTIONS TO THE COMMAND LINE TOOL 17
Format name Standard Read Write File type File extension Reference
OPF No Yes Yes ASCII/UTF8 opf Appendix D
MPS Yes Yes Yes ASCII mps Appendix B
LP Partially Yes Yes ASCII lp Appendix C
OSiL XML Yes No Yes ASCII/UTF8 xml Appendix E
Binary task format No Yes Yes Binary mbt
Table 4.1: Supported le formats.
4.3 Passing options to the command line tool
It is possible to modify the behavior of MOSEK by setting appropriate parameters. E.g assume that
a linear optimization problem should be solved with the primal simplex optimizer rather than the
default interior-point optimizer. This is done by setting the parameter MSK IPAR OPTIMIZER to the
value MSK OPTIMIZER PRIMAL SIMPLEX. To accomplish this append
-d MSK_IPAR_OPTIMIZER MSK_OPTIMIZER_PRIMAL_SIMPLEX.
to the command line. E.g the command
mosek -d MSK_IPAR_OPTIMIZER MSK_OPTIMIZER_PRIMAL_SIMPLEX lo1.opf
solves the problem specied by lo1.opf using the primal simplex optimizer. For further information
on the parameters available in MOSEK please see Appendix H.
4.4 Reading and writing problem data les
MOSEK reads and writes problem data les in the formats presented in Table 4.1. The columns of
Table 4.1 show:
The name of the format.
Whether the format is an industry standard format.
If the format can be read by MOSEK.
If the format can be written by MOSEK.
The generic le type of the format, i.e. ASCII, UTF8, or binary.
The le extension for the format
The location of information about the format.
4.4.1 Reading compressed data les
MOSEK can read and write data les compressed with gzip
1
For mosek to recognize a le as a gzip compressed le it must have the extension .gz. E.g the
command
mosek myfile.mps.gz
will automatically decompress the le while reading it.
4.4.2 Converting from one format and to another
It is possible to use MOSEK to convert a problem le from one format to another. For instance assume
the MPS le myprob.mps should be converted to an LP le named myprob.lp. This can be achieved
with the command
mosek myprob.mps -out myprob.lp -x
Converting an MPS le to a LP le and back to an MPS le permutes the rows and columns of
the original problem; this has no inuence on the problem, but variables and constraints may appear
in a dierent order.
4.5 Hot-start
Often a sequence of closely related optimization problems has to be solved. In such a case it can
be expected that a previous optimal solution can serve as a good starting point when the modied
problem is optimized.
Currently, only the simplex optimizer and the mixed-integer optimizer in MOSEK can exploit a
guess for the optimal solution. The simplex optimizer can exploit an arbitrary guess for the optimal
solution whereas the mixed-integer optimizer requires a feasible integer solution. For both the simplex
optimizer and the mixed-integer optimizer it holds that if the guess is good then the optimizer may
be able to reduce the solution time signicantly when exploiting the guess.
4.5.1 An example
Assume that the example
minimize c
1
x
1
9x
2
,
subject to 7/10x
1
+ 1x
2
630,
1/2x
1
+ 5/6x
2
600,
1x
1
+ 2/3x
2
708,
1/10x
1
+ 1/4x
2
135,
x
1
, x
2
0.
(4.8)
should be solved for c
1
identical to 5 and 10. Clearly, a solution for one c
1
value will also be feasible
for another value. Therefore, it might be worthwhile to exploit the previous optimal solution when
reoptimizing the problem.
1
gzip is a public domain compression format. For further details about gzip consult http://www.gzip.org
4.6. FURTHER INFORMATION 19
Assume that two MPS les have been created each corresponding to one of the c
1
values then the
commands
2
mosek lo1.mps -baso .\lo1.bas
mosek lo1-b.mps -basi .\lo1.bas -baso .\lo1-b.bas
-d MSK_IPAR_OPTIMIZER MSK_OPTIMIZER_PRIMAL_SIMPLEX
demonstrates how to exploit the previous optimal solution in the second optimization.
In the rst line MOSEK optimizes the rst version of the optimization problem where c
1
is identical
to 10. The -baso .\lo1.bas command line option makes sure that the optimal basic solution is
written to the le .\lo1.bas.
In the second line the second instance of the problem is optimized. The -basi .\lo1.bas command
line option forces MOSEK to read the previous optimal solution which MOSEK will try to exploit
automatically. The -baso .\lo1-b.bas command line option makes sure that the optimal basic
solution is written to the .\lo1-b.bas le. Finally, the
-d MSK_IPAR_OPTIMIZER MSK_OPTIMIZER_PRIMAL_SIMPLEX
command line option makes sure that the primal simplex optimizer is used for the reoptimization. This
is important because the interior-point optimizer used by default does not exploit a previous optimal
solution.
4.6 Further information
Additional information about the MOSEK command line tool is available in Appendix A.
4.7 Solution le ltering
The MOSEK solution les can be very space consuming for large problems. One way to cut down the
solution le size is only to include variables which optimal value is in a certain interesting range i.e
[0.01, 0.99]. This can be done by setting the MOSEK parameters
MSK_SPAR_SOL_FILTER_XX_LOW 0.01
MSK_SPAR_SOL_FILTER_XX_UPR 0.99
For further details consult the parameters MSK SPAR SOL FILTER XC LOW and MSK SPAR SOL FILTER XC UPR.
2
The second line should not be broken into two separate lines.
Chapter 5
MOSEK and AMPL
AMPL is a modeling language for specifying linear and nonlinear optimization models in a natural
way. AMPL also makes it easy to solve the problem and e.g. display the solution or part of it.
We will not discuss the specics of the AMPL language here but instead refer the reader to [13]
and the AMPL website http://www.ampl.com.
AMPL cannot solve optimization problems by itself but requires a link to an appropriate optimizer
such as MOSEK. The MOSEK distribution includes an AMPL link which makes it possible to use
MOSEK as an optimizer within AMPL.
5.1 Invoking the AMPL shell
The MOSEK distribution by default comes with the AMPL shell installed. To invoke the AMPL shell
type:
mampl
5.2 Applicability
It is possible to specify problems in AMPL that cannot be solved by MOSEK. The optimization
problem must be a smooth convex optimization problem as discussed in Section 9.4.
5.3 An example
In many instances, you can successfully apply MOSEK simply by specifying the model and data,
setting the solver option to MOSEK, and typing solve. First to invoke the AMPL shell type:
mampl
when the AMPL shell has started type the commands:
ampl: model diet.mod;
ampl: data diet.dat;
21
22 CHAPTER 5. MOSEK AND AMPL
Value Message
0 the solution is optimal.
100 suboptimal primal solution.
101 superoptimal (dual feasible) solution.
150 the solution is near optimal.
200 primal infeasible problem.
300 dual infeasible problem.
400 too many iterations.
500 solution status is unknown.
501 ill-posed problem, solution status is unknown.
501 The value - 501 is a MOSEK response code.
See Appendix I.40 for all MOSEK response codes.
Table 5.1: Interpretation of solve result num.
ampl: option solver mosek;
ampl: solve;
The resulting output is:
MOSEK finished.
Problem status - PRIMAL_AND_DUAL_FEASIBLE
Solution status - OPTIMAL
Primal objective - 14.8557377
Dual objective - 14.8557377
Objective = Total_Cost
5.4 Determining the outcome of an optimization
The AMPL parameter solve result num is used to indicate the outcome of the optimization process.
It is used as follows
ampl: display solve_result_num
Please refer to table 5.1 for possible values of this parameter.
5.5 Optimizer options
5.5.1 The MOSEK parameter database
The MOSEK optimizer has options and parameters controlling such things as the termination criterion
and which optimizer is used. These parameters can be modied within AMPL as shown in the example
below:
5.6. CONSTRAINT AND VARIABLE NAMES 23
ampl: option mosek_options
ampl? msk_ipar_optimizer = msk_optimizer_primal_simplex \
ampl? msk_ipar_sim_max_iterations = 100000;
ampl: solve;
In the example above a string called mosek options is created which contains the parameter settings.
Each parameter setting has the format
parameter name = value
where parameter name can be any valid MOSEK parameter name. See Appendix H for a description
of all valid MOSEK parameters.
An alternative way of specifying the options is
ampl? msk_ipar_optimizer = msk_optimizer_primal_simplex
ampl? msk_ipar_sim_max_iterations = 100000;
New options can also be appended to an existing option string as shown below
ampl: option mosek_options $mosek_options
ampl? msk_ipar_sim_print_freq = 0 msk_ipar_sim_max_iterations = 1000;
The expression $mosek_options expands to the current value of the option. Line two in the example
appends an additional value msk ipar sim max iterations to the option string.
5.5.2 Options
5.5.2.1 outlev
MOSEK also recognizes the outlev option which controls the amount of printed output. 0 means no
printed output and a higher value means more printed output. An example of setting outlev is as
follows:
ampl: option mosek_options outlev=2;
5.5.2.2 wantsol
MOSEK recognize the option wantsol. We refer the reader to the AMPL manual [13] for details about
this option.
5.6 Constraint and variable names
AMPL assigns meaningfull names to all the constraints and variables. Since MOSEK uses item names
in error and log messages, it may be useful to pass the AMPL names to MOSEK. Using the command
ampl: option mosek_auxfiles rc;
before the
solve;
command makes MOSEK obtain the constraint and variable names automatically.
5.7 Which solution is returned to AMPL
The MOSEK optimizer can produce three types of solutions: basic, integer, and interior point solutions.
For nonlinear problems only an interior solution is available. For linear optimization problems opti-
mized by the interior-point optimizer with basis identication turned on both a basic and an interior
point solution are calculated. The simplex algorithm produces only a basic solution. Whenever both
an interior and a basic solution are available, the basic solution is returned. For problems containing
integer variables, the integer solution is returned to AMPL.
5.8 Hot-start
Frequently, a sequence of optimization problems is solved where each problem diers only slightly from
the previous problem. In that case it may be advantageous to use the previous optimal solution to
hot-start the optimizer. Such a facility is available in MOSEK only when the simplex optimizer is
used.
The hot-start facility exploits the AMPL variable sux sstatus to communicate the optimal basis
back to AMPL, and AMPL uses this facility to communicate an initial basis to MOSEK. The following
example demonstrates this feature.
ampl? msk_ipar_optimizer = msk_optimizer_primal_simplex outlev=2;
ampl: solve;
ampl: display Buy.sstatus;
ampl: solve;
Accepted: msk_ipar_optimizer = MSK_OPTIMIZER_PRIMAL_SIMPLEX
Accepted: outlev = 2
Computer - Platform : Linux/64-X86
Computer - CPU type : Intel-P4
MOSEK - task name :
MOSEK - objective sense : min
MOSEK - problem type : LO (linear optimization problem)
MOSEK - constraints : 7 variables : 9
MOSEK - integer variables : 0
Optimizer started.
Simplex optimizer started.
Presolve started.
Linear dependency checker started.
Linear dependency checker terminated.
Presolve - Stk. size (kb) : 0
Eliminator - tries : 0 time : 0.00
Eliminator - elims : 0
Lin. dep. - tries : 1 time : 0.00
5.8. HOT-START 25
Lin. dep. - number : 0
Presolve terminated. Time: 0.00
Primal simplex optimizer started.
Primal simplex optimizer setup started.
Primal simplex optimizer setup terminated.
Optimizer - solved problem : the primal
Optimizer - constraints : 7 variables : 9
Optimizer - hotstart : no
ITER DEGITER(%) PFEAS DFEAS POBJ DOBJ TIME TOTTIME
0 0.00 1.40e+03 NA 1.2586666667e+01 NA 0.00 0.01
3 0.00 0.00e+00 NA 1.4855737705e+01 NA 0.00 0.01
Primal simplex optimizer terminated.
Simplex optimizer terminated. Time: 0.00.
Optimizer terminated. Time: 0.01
Return code - 0 [MSK_RES_OK]
MOSEK finished.
Problem status : PRIMAL_AND_DUAL_FEASIBLE
Solution status : OPTIMAL
Primal objective : 14.8557377
Dual objective : 14.8557377
Buy.sstatus [*] :=
Quarter Pounder w/ Cheese bas
McLean Deluxe w/ Cheese low
Big Mac low
Filet-O-Fish low
McGrilled Chicken low
Fries, small bas
Sausage McMuffin low
1% Lowfat Milk bas
Orange Juice low
;
Accepted: msk_ipar_optimizer = MSK_OPTIMIZER_PRIMAL_SIMPLEX
Accepted: outlev = 2
Basic solution
Problem status : UNKNOWN
Solution status : UNKNOWN
Primal - objective: 1.4855737705e+01 eq. infeas.: 3.97e+03 max bound infeas.: 2.00e+03
Dual - objective: 0.0000000000e+00 eq. infeas.: 7.14e-01 max bound infeas.: 0.00e+00
Computer - Platform : Linux/64-X86
Computer - CPU type : Intel-P4
MOSEK - task name :
MOSEK - objective sense : min
MOSEK - problem type : LO (linear optimization problem)
MOSEK - constraints : 7 variables : 9
MOSEK - integer variables : 0
Optimizer started.
Simplex optimizer started.
Presolve started.
Presolve - Stk. size (kb) : 0
Eliminator - tries : 0 time : 0.00
Eliminator - elims : 0
Lin. dep. - tries : 0 time : 0.00
Lin. dep. - number : 0
Presolve terminated. Time: 0.00
Primal simplex optimizer started.
Primal simplex optimizer setup started.
Primal simplex optimizer setup terminated.
Optimizer - solved problem : the primal
Optimizer - hotstart : yes
Optimizer - Num. basic : 7 Basis rank : 7
Optimizer - Valid bas. fac. : no
ITER DEGITER(%) PFEAS DFEAS POBJ DOBJ TIME TOTTIME
0 0.00 0.00e+00 NA 1.4855737705e+01 NA 0.00 0.01
0 0.00 0.00e+00 NA 1.4855737705e+01 NA 0.00 0.01
Primal simplex optimizer terminated.
Simplex optimizer terminated. Time: 0.00.
Optimizer terminated. Time: 0.01
MOSEK finished.
Please note that the second solve takes fewer iterations since the previous optimal basis is reused.
5.9 Sensitivity analysis
MOSEK can calculate sensitivity information for the objective and constraints. To enable sensitivity
information set the option:
sensitivity = 1
Results are returned in variable/constraint suxes as follows:
.down Smallest value of objective coecient/right hand side before the optimal basis changes.
.up Largest value of objective coecient/right hand side before the optimal basis changes.
.current Current value of objective coecient/right hand side.
For ranged constraints sensitivity information is returned only for the lower bound.
The example below returns sensitivity information on the diet model.
5.9. SENSITIVITY ANALYSIS 27
ampl: option mosek_options sensitivity=1;
ampl: solve;
#display sensitivity information and current solution.
ampl: display _var.down,_var.current,_var.up,_var;
#display sensitivity information and optimal dual values.
ampl: display _con.down,_con.current,_con.up,_con;
MOSEK finished.
suffix up OUT;
suffix down OUT;
suffix current OUT;
: _var.down _var.current _var.up _var :=
1 1.37385 1.84 1.86075 4.38525
2 1.8677 2.19 Infinity 0
3 1.82085 1.84 Infinity 0
4 1.35466 1.44 Infinity 0
5 1.57633 2.29 Infinity 0
6 0.094 0.77 0.794851 6.14754
7 1.22759 1.29 Infinity 0
8 0.57559 0.6 0.910769 3.42213
9 0.657279 0.72 Infinity 0
;
ampl: display _con.down,_con.current,_con.up,_con;
: _con.down _con.current _con.up _con :=
1 -Infinity 2000 3965.37 0
2 297.6 350 375 0.0277049
3 -Infinity 55 172.029 0
4 63.0531 100 195.388 0.0267541
5 -Infinity 100 132.213 0
6 -Infinity 100 234.221 0
7 17.6923 100 142.821 0.0248361
;
5.10 Using the command line version of the AMPL interface
AMPL can generate a data le containing all the optimization problem and all relevant information
which can then be read and solved by the MOSEK command line tool.
When the problem has been loaded into AMPL, the commands
ampl: option auxfiles rc;
ampl: write bprob;
will make AMPL write the appropriate data les, i.e.
prob.nl
prob.col
prob.row
Then the problem can be solved using the command line version of MOSEK as follows
mosek prob.nl outlev=10 -a
The -a command line option indicates that MOSEK is invoked in AMPL mode. When MOSEK is
invoked in AMPL mode the normal MOSEK command line options should appear after the -a option
except for the le name which should be the rst argument. As the above example demonstrates
MOSEK accepts command line options as specied by the AMPL convention. Which command line
arguments MOSEK accepts in AMPL mode can be viewed by executing
mosek -= -a
For linear, quadratic and quadratic constrained problems a text le representation of the problem
can be obtained using one of the commands
mosek prob.nl -a -x -out prob.mps
mosek prob.nl -a -x -out prob.opf
mosek prob.nl -a -x -out prob.lp
Chapter 6
MOSEK and GAMS
It is possible to call MOSEK from the GAMS modeling language . In order to do so a special
GAMS/MOSEK link must be obtained from the GAMS Corporation.
29
30 CHAPTER 6. MOSEK AND GAMS
Chapter 7
MOSEK and MATLAB
The MOSEK optimization toolbox for MATLAB is an easy to use interface to MOSEK that makes it
possible to use MOSEK from within MATLAB.
The optimization toolbox is included in the MOSEK optimization tools distribution. See the
separate documentation for the MATLAB toolbox for details.
31
32 CHAPTER 7. MOSEK AND MATLAB
Chapter 8
Interfaces to MOSEK
8.1 The optimizer API
The MOSEK optimizer API is an ecient interface to the optimizers implemented in MOSEK. E.g the
interface makes it possible to call the linear optimizer from a C++ or Java program. The optimizer
API is available for the languages
C/C++/Delphi.
Java.
.NET (Visual Basic, C#, Managed C++, etc).
Python.
Further details about the optimizer APIs are available at
mosek\6\help\index.html
or online at
http://www.mosek.com/documentation/
33
34 CHAPTER 8. INTERFACES TO MOSEK
Chapter 9
Modelling
In this chapter we will discuss the following issues:
The formal denitions of the problem types that MOSEK can solve.
The solution information produced by MOSEK.
The information produced by MOSEK if the problem is infeasible.
A set of examples showing dierent ways of formulating commonly occurring problems so that
they can be solved by MOSEK.
Recommendations for formulating optimization problems.
9.1 Linear optimization
A linear optimization problem can be written as
minimize c
T
x + c
f
subject to l
c
Ax u
c
,
l
x
x u
x
,
(9.1)
where
m is the number of constraints.
n is the number of decision variables.
x R
n
is a vector of decision variables.
c R
n
is the linear part of the objective function.
A R
mn
is the constraint matrix.
l
c
R
m
is the lower limit on the activity for the constraints.
u
c
R
m
is the upper limit on the activity for the constraints.
35
36 CHAPTER 9. MODELLING
l
x
R
n
is the lower limit on the activity for the variables.
u
x
R
n
is the upper limit on the activity for the variables.
A primal solution (x) is (primal) feasible if it satises all constraints in (9.1). If (9.1) has at least
one primal feasible solution, then (9.1) is said to be (primal) feasible.
In case (9.1) does not have a feasible solution, the problem is said to be (primal) infeasible.
9.1.1 Duality for linear optimization
Corresponding to the primal problem (9.1), there is a dual problem
maximize (l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+(l
x
)
T
s
x
l
(u
x
)
T
s
x
u
+ c
f
subject to A
T
y + s
x
l
s
x
u
= c,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
(9.2)
If a bound in the primal problem is plus or minus innity, the corresponding dual variable is xed at
0, and we use the convention that the product of the bound value and the corresponding dual variable
is 0. E.g.
l
x
j
= (s
x
l
)
j
= 0 and l
x
j
(s
x
l
)
j
= 0.
This is equivalent to removing variable (s
x
l
)
j
from the dual problem.
A solution
(y, s
c
l
, s
c
u
, s
x
l
, s
x
u
)
to the dual problem is feasible if it satises all the constraints in (9.2). If (9.2) has at least one feasible
solution, then (9.2) is (dual) feasible, otherwise the problem is (dual) infeasible.
We will denote a solution
(x, y, s
c
l
, s
c
u
, s
x
l
, s
x
u
)
so that x is a solution to the primal problem (9.1), and
(y, s
c
l
, s
c
u
, s
x
l
, s
x
u
)
is a solution to the corresponding dual problem (9.2). A solution which is both primal and dual feasible
is denoted a primal-dual feasible solution.
9.1.1.1 A primal-dual feasible solution
Let
(x
, y
, (s
c
l
)
, (s
c
u
)
, (s
x
l
)
, (s
x
u
)
)
be a primal-dual feasible solution, and let
(x
c
)
:= Ax
.
9.1. LINEAR OPTIMIZATION 37
For a primal-dual feasible solution we dene the optimality gap as the dierence between the primal
and the dual objective value,
c
T
x
+ c
f
((l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
+ c
f
)
=
m
i=1
((s
c
l
)
i
((x
c
i
)
l
c
i
) + (s
c
u
)
i
(u
c
i
(x
c
i
)
) +
n
j=1
((s
x
l
)
j
(x
j
l
x
j
) + (s
x
u
)
j
(u
x
j
x
j
))
0
where the rst relation can be obtained by multiplying the dual constraints (9.2) by x and x
c
respec-
tively, and the second relation comes from the fact that each term in each sum is nonnegative. It
follows that the primal objective will always be greater than or equal to the dual objective.
We then dene the duality gap as the dierence between the primal objective value and the dual
objective value, i.e.
c
T
x
+ c
f
((l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
+ c
f
)
Please note that the duality gap will always be nonnegative.
9.1.1.2 An optimal solution
It is well-known that a linear optimization problem has an optimal solution if and only if there exist
feasible primal and dual solutions so that the duality gap is zero, or, equivalently, that the comple-
mentarity conditions
(s
c
l
)
i
((x
c
i
)
l
c
i
) = 0, i = 1, . . . , m,
(s
c
u
)
i
(u
c
i
(x
c
i
)
) = 0, i = 1, . . . , m,
(s
x
l
)
j
(x
j
l
x
j
) = 0, j = 1, . . . , n,
(s
x
u
)
j
(u
x
j
x
j
) = 0, j = 1, . . . , n
are satised.
If (9.1) has an optimal solution and MOSEK solves the problem successfully, both the primal and
dual solution are reported, including a status indicating the exact state of the solution.
9.1.1.3 Primal infeasible problems
If the problem (9.1) is infeasible (has no feasible solution), MOSEK will report a certicate of primal
infeasibility: The dual solution reported is a certicate of infeasibility, and the primal solution is
undened.
A certicate of primal infeasibility is a feasible solution to the modied dual problem
maximize (l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
subject to A
T
y + s
x
l
s
x
u
= 0,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
(9.3)
so that the objective is strictly positive, i.e. a solution
(y
, (s
c
l
)
, (s
c
u
)
, (s
x
l
)
, (s
x
u
)
)
to (9.3) so that
(l
c
)
T
(s
c
l
)
(u
c
)
T
(s
c
u
)
+ (l
x
)
T
(s
x
l
)
(u
x
)
T
(s
x
u
)
> 0.
Such a solution implies that (9.3) is unbounded, and that its dual is infeasible.
We note that the dual of (9.3) is a problem which constraints are identical to the constraints of
the original primal problem (9.1): If the dual of (9.3) is infeasible, so is the original primal problem.
9.1.1.4 Dual infeasible problems
If the problem (9.2) is infeasible (has no feasible solution), MOSEK will report a certicate of dual
infeasibility: The primal solution reported is a certicate of infeasibility, and the dual solution is
undened.
A certicate of dual infeasibility is a feasible solution to the problem
minimize c
T
x
subject to Ax x
c
= 0,
l
c
x
c
u
c
,
l
x
x u
x
(9.4)
where
l
c
i
=
_
0, if l
c
i
> ,
otherwise
and u
c
i
:=
_
0, if u
c
i
< ,
otherwise
and
l
x
j
=
_
0, if l
x
j
> ,
otherwise
and u
x
j
:=
_
0, if u
x
j
< ,
otherwise
so that the objective value c
T
x is negative. Such a solution implies that (9.4) is unbounded, and that
the dual of (9.4) is infeasible.
We note that the dual of (9.4) is a problem which constraints are identical to the constraints of
the original dual problem (9.2): If the dual of (9.4) is infeasible, so is the original dual problem.
9.1.2 Primal and dual infeasible case
In case that both the primal problem (9.1) and the dual problem (9.2) are infeasible, MOSEK will
report only one of the two possible certicates which one is not dened (MOSEK returns the rst
certicate found).
9.2 Quadratic and quadratically constrained optimization
A convex quadratic optimization problem is an optimization problem of the form
minimize
1
2
x
T
Q
o
x + c
T
x + c
f
subject to l
c
k

1
2
x
T
Q
k
x +
n1
j=0
a
k,i
x
j
u
c
k
, k = 0, . . . , m1,
l
x
x u
x
, j = 0, . . . , n 1,
(9.5)
where the convexity requirement implies that
Q
o
is a symmetric positive semi-denite matrix.
If l
c
k
= , then Q
k
is a symmetric positive semi-denite matrix.
9.2. QUADRATIC AND QUADRATICALLY CONSTRAINED OPTIMIZATION 39
If u
c
k
= , then Q
k
is a symmetric negative semi-denite matrix.
If l
k
> and u
k
k
< , then Q
k
is a zero matrix.
The convexity requirement is very important and it is strongly recommended that MOSEK is
applied to convex problems only.
9.2.1 A general recommendation
Any convex quadratic optimization problem can be reformulated as a conic optimization problem.
It is our experience that for the majority of practical applications it is better to cast them as conic
problems because
the resulting problem is convex by construction, and
the conic optimizer is more ecient than the optimizer for general quadratic problems.
See Section 9.3.3.1 for further details.
9.2.2 Reformulating as a separable quadratic problem
The simplest quadratic optimization problem is
minimize 1/2x
T
Qx + c
T
x
subject to Ax = b,
x 0.
(9.6)
The problem (9.6) is said to be a separable problem if Q is a diagonal matrix or, in other words, if the
quadratic terms in the objective all have this form
x
2
j
instead of this form
x
j
x
i
.
The separable form has the following advantages:
It is very easy to check the convexity assumption, and
the simpler structure in a separable problem usually makes it easier to solve.
It is well-known that a positive semi-denite matrix Q can always be factorized, i.e. a matrix F
exists so that
Q = F
T
F. (9.7)
In many practical applications of quadratic optimization F is known explicitly; e.g. if Q is a covariance
matrix, F is the set of observations producing it.
Using (9.7), the problem (9.6) can be reformulated as
minimize 1/2y
T
Iy + c
T
x
subject to Ax = b,
Fx y = 0,
x 0.
(9.8)
The problem (9.8) is also a quadratic optimization problem and has more constraints and variables
than (9.6). However, the problem is separable. Normally, if F has fewer rows than columns, it is
worthwhile to reformulate as a separable problem. Indeed consider the extreme case where F has one
dense row and hence Q will be a dense matrix.
The idea presented above is applicable to quadratic constraints too. Now, consider the constraint
1/2x
T
(F
T
F)x b (9.9)
where F is a matrix and b is a scalar. (9.9) can be reformulated as
1/2y
T
Iy b,
Fx y = 0.
It should be obvious how to generalize this idea to make any convex quadratic problem separable.
Next, consider the constraint
1/2x
T
(D + F
T
F)x b
where D is a positive semi-denite matrix, F is a matrix, and b is a scalar. We assume that D has a
simple structure, e.g. that D is a diagonal or a block diagonal matrix. If this is the case, it may be
worthwhile performing the reformulation
1/2((x
T
Dx) + y
T
Iy) b,
Fx y = 0.
Now, the question may arise: When should a quadratic problem be reformulated to make it sepa-
rable or near separable? The simplest rule of thumb is that it should be reformulated if the number
of non-zeros used to represent the problem decreases when reformulating the problem.
9.3 Conic optimization
Conic optimization can be seen as a generalization of linear optimization. Indeed a conic optimization
problem is a linear optimization problem plus a constraint of the form
x (
where ( is a convex cone. A complete conic problem has the form
minimize c
T
x + c
f
subject to l
c
Ax u
c
,
l
x
x u
x
,
x (.
(9.10)
The cone ( can be a Cartesian product of p convex cones, i.e.
( = (
1
(
p
in which case x ( can be written as
x = (x
1
, . . . , x
p
), x
1
(
1
, . . . , x
p
(
p
9.3. CONIC OPTIMIZATION 41
where each x
t
R
nt
. Please note that the n-dimensional Euclidean space R
n
is a cone itself, so simple
linear variables are still allowed.
MOSEK supports only a limited number of cones, specically
( = (
1
(
p
where each (
t
has one of the following forms
R set:
(
t
= x R
n
t
.
Quadratic cone:
(
t
=
_
_
_
x R
n
t
: x
1

_
n
t
j=2
x
2
j
_
_
_
.
Rotated quadratic cone:
(
t
=
_
_
_
x R
n
t
: 2x
1
x
2

n
t
j=3
x
2
j
, x
1
, x
2
0
_
_
_
.
Although these cones may seem to provide only limited expressive power they can be used to model
a large range of problems as demonstrated in Section 9.3.3.
9.3.1 Duality for conic optimization
The dual problem corresponding to the conic optimization problem (9.10) is given by
maximize (l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+(l
x
)
T
s
x
l
(u
x
)
T
s
x
u
+ c
f
subject to A
T
y + s
x
l
s
x
u
+ s
x
n
= c,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0,
s
x
n
(
(9.11)
where the dual cone (
is a product of the cones

(
= (
1
(
p
where each (
t
is the dual cone of (
t
. For the cone types MOSEK can handle, the relation between the
primal and dual cone is given as follows:
R set:
(
t
=
_
x R
n
t
_
(
t
:=
_
s R
n
t
: s = 0
_
.
Quadratic cone:
(
t
:=
_
_
_
x R
n
t
: x
1

_
n
t
j=2
x
2
j
_
_
_
(
t
= (
t
.
(
t
:=
_
_
_
x R
n
t
: 2x
1
x
2

n
t
j=3
x
2
j
, x
1
, x
2
0
_
_
_
. (
t
= (
t
.
Please note that the dual problem of the dual problem is identical to the original primal problem.
9.3.2 Infeasibility
In case MOSEK nds a problem to be infeasible it reports a certicate of the infeasibility. This works
exactly as for linear problems (see Sections 9.1.1.3 and 9.1.1.4).
9.3.3 Examples
This section contains several examples of inequalities and problems that can be cast as conic optimiza-
tion problems.
9.3.3.1 Quadratic objective and constraints
From Section 9.2.2 we know that any convex quadratic problem can be stated on the form
minimize 0.5 |Fx|
2
+ c
T
x,
subject to 0.5 |Gx|
2
+ a
T
x b,
(9.12)
where F and G are matrices and c and a are vectors. For simplicity we assume that there is only
one constraint, but it should be obvious how to generalize the methods to an arbitrary number of
constraints.
Problem (9.12) can be reformulated as
minimize 0.5 |t|
2
+ c
T
x,
subject to 0.5 |z|
2
+ a
T
x b,
Fx t = 0,
Gx z = 0
(9.13)
after the introduction of the new variables t and z. It is easy to convert this problem to a conic
quadratic optimization problem, i.e.
minimize v + c
T
x,
subject to p + a
T
x = b,
Fx t = 0,
Gx z = 0,
w = 1,
q = 1,
|t|
2
2vw, v, w 0,
|z|
2
2pq, p, q 0.
(9.14)
In this case we can model the last two inequalities using rotated quadratic cones.
If we assume that F is a non-singular matrix e.g. a diagonal matrix then
x = F
1
t
and hence we can eliminate x from the problem to obtain:
minimize v + c
T
F
1
t,
subject to p + a
T
F
1
t = b,
GF
1
t z = 0,
w = 1,
q = 1,
|t|
2
2vw, v, w 0,
|z|
2
2pq, p, q 0.
(9.15)
In most cases MOSEK performs this reduction automatically during the presolve phase before the
optimization is performed.
9.3.3.2 Minimizing a sum of norms
The next example is the problem of minimizing a sum of norms, i.e. the problem
minimize
k
i=1
_
_
x
i
_
_
subject to Ax = b,
(9.16)
where
x :=
_
_
x
1
.
.
.
x
k
_
_.
This problem is equivalent to
minimize
k
i=1
z
i
subject to Ax = b,
_
_
x
i
_
_
z
i
, i = 1, . . . , k,
(9.17)
which in turn is equivalent to
minimize
k
i=1
z
i
subject to Ax = b,
(z
i
, x
i
) (
i
, i = 1, . . . , k
(9.18)
where all (
i
are of the quadratic type, i.e.
(
i
:=
_
(z
i
, x
i
) : z
i

_
_
x
i
_
_
_
.
The dual problem corresponding to (9.18) is
maximize b
T
y
subject to A
T
y + s = c,
t
i
= 1, i = 1, . . . , k,
(t
i
, s
i
) (
i
, i = 1, . . . , k
(9.19)
where
s :=
_
_
s
1
.
.
.
s
k
_
_.
This problem is equivalent to
maximize b
T
y
subject to A
T
y + s = c,
_
_
s
i
_
_
2
2
1, i = 1, . . . , k.
(9.20)
Please note that in this case the dual problem can be reduced to an ordinary convex quadratically
constrained optimization problem due to the special structure of the primal problem. In some cases
it turns out that it is much better to solve the dual problem (9.19) rather than the primal problem
(9.18).
9.3.3.3 Modelling polynomial terms using conic optimization
Generally an arbitrary polynomial term of the form
fx
g
cannot be represented with conic quadratic constraints, however in the following we will demonstrate
some special cases where it is possible.
A particular simple polynomial term is the reciprocal, i.e.
1
x
.
Now, a constraint of the form
1
x
y
where it is required that x > 0 is equivalent to
1 xy and x > 0
which in turn is equivalent to
z =

2,
z
2
2xy.
The last formulation is a conic constraint plus a simple linear equality.
E.g., consider the problem
minimize c
T
x
subject to
n
j=1
fj
xj
b,
x 0,
where it is assumed that f
j
> 0 and b > 0. This problem is equivalent to
minimize c
T
x
subject to
n
j=1
f
j
z
j
= b,
v
j
=

2, j = 1, . . . , n,
v
2
j
2z
j
x
j
, j = 1, . . . , n,
x, z 0,
(9.21)
because
v
2
j
= 2 2z
j
x
j
implies that
1
x
j
z
j
and
n
j=1
f
j
x
j
j=1
f
j
z
j
= b.
The problem (9.21) is a conic quadratic optimization problem having n 3-dimensional rotated
quadratic cones.
The next example is the constraint
x [t[,
x 0,
where both t and x are variables. This set is identical to the set
t
2
2xz,
z = 0.5,
x, z, 0.
(9.22)
Occasionally, when modeling the market impact term in portfolio optimization, the polynomial
term x
3
2
occurs. Therefore, consider the set dened by the inequalities
x
1.5
t,
0 x.
(9.23)
We will exploit that x
1.5
= x
2
/
x . First dene the set

x
2
2st,
s, t 0.
(9.24)
Now, if we can make sure that
2s
x,
then we have the desired result since this implies that
x
1.5
=
x
2
x

x
2
2s
t.
Please note that s can be chosen freely and that

x = 2s is a valid choice.
Let
x
2
2st,
w
2
2vr,
x = v,
s = w,
r =
1
8
,
s, t, v, r 0,
(9.25)
then
s
2
= w
2
2vr
=
v
4
=
x
4
.
Moreover,
x
2
2st,
2
_
x
4
t
leading to the conclusion that
x
1.5
t.
(9.25) is a conic reformulation which is equivalent to (9.23). Please note that the x 0 constraint
does not appear explicitly in (9.24) and (9.25), but implicitly since x = v 0.
Finally, it should be mentioned that any polynomial term of the form x
g
where g is a positive
rational number can be represented using conic quadratic constraints [2, pp. 12-13]
9.3.3.4 Further reading
If you want to learn more about what can be modeled as a conic optimization problem we recommend
the references [2, 11, 16].
9.3.4 Potential pitfalls in conic optimization
While a linear optimization problem either has a bounded optimal solution or is infeasible, the conic
case is not as simple as that.
9.3.4.1 Non-attainment in the primal problem
Consider the example
minimize z
subject to 2yz x
2
,
x =

2,
y, z 0,
(9.26)
which corresponds to the problem
minimize
1
y
subject to y 0.
(9.27)
Clearly, the optimal objective value is zero but it is never attained because implicitly we assume that
the optimal y is nite.
9.3.4.2 Non-attainment in the dual problem
Next, consider the example
minimize x
4
subject to x
3
+ x
4
= 1,
x
1
= 0,
x
2
= 1,
2x
1
x
2
x
2
3
,
x
1
, x
2
0,
(9.28)
which has the optimal solution
x
1
= 0, x
2
= 1, x
3
= 0 and x
4
= 1
implying that the optimal primal objective value is 1.
Now, the dual problem corresponding to (9.28) is
maximize y
1
+ y
3
subject to y
2
+ s
1
= 0,
y
3
+ s
2
= 0,
y
1
+ s
3
= 0,
y
1
= 1,
2s
1
s
2
s
2
3
,
s
1
, s
2
0.
(9.29)
Therefore,
y
1
= 1
and
s
3
= 1.
This implies that
2s
1
s
2
(s
3
)
2
= 1
and hence s
2
> 0. Given this fact we can conclude that
y
1
+ y
3
= 1 s
2
< 1
implying that the optimal dual objective value is 1, however, this is never attained. Hence, no primal-
dual bounded optimal solution with zero duality gap exists. Of course it is possible to nd a primal-dual
feasible solution such that the duality gap is close to zero, but then s
1
will be similarly large. This is
likely to make the problem (9.28) hard to solve.
An inspection of the problem (9.28) reveals the constraint x
1
= 0, which implies that x
3
= 0. If
we either add the redundant constraint
x
3
= 0
to the problem (9.28) or eliminate x
1
and x
3
from the problem it becomes easy to solve.
9.4 Nonlinear convex optimization
MOSEK is capable of solving smooth (twice dierentiable) convex nonlinear optimization problems of
the form
minimize f(x) + c
T
x
subject to g(x) + Ax x
c
= 0,
l
c
x
c
u
c
,
l
x
x u
x
,
(9.30)
where
m is the number of constraints.
n is the number of decision variables.
x R
n
is a vector of decision variables.
x
c
R
m
is a vector of constraints or slack variables.
c R
n
is the linear part objective function.
A R
mn
l
c
R
m
u
c
R
m
l
x
R
n
u
x
R
n
f : R
n
R is a nonlinear function.
g : R
n
R
m
is a nonlinear vector function.
This means that the ith constraint has the form
l
c
i
g
i
(x) +
n
j=1
a
i,j
x
j
u
c
i
when the x
c
i
variable has been eliminated.
The linear term Ax is not included in g(x) since it can be handled much more eciently as a
separate entity when optimizing.
The nonlinear functions f and g must be smooth in all x [l
x
; u
x
]. Moreover, f(x) must be a
convex function and g
i
(x) must satisfy
l
c
i
= g
i
(x) is convex,
u
c
i
= g
i
(x) is concave,
< l
c
i
u
c
i
< g
i
(x) = 0.
9.5. RECOMMENDATIONS 49
9.4.1 Duality
So far, we have not discussed what happens when MOSEK is used to solve a primal or dual infeasible
problem. In the following section these issues are addressed.
Similar to the linear case, MOSEK reports dual information in the general nonlinear case. Indeed
in this case the Lagrange function is dened by
L(x
c
, x, y, s
c
l
, s
c
u
, s
x
l
, s
x
u
) := f(x) + c
T
x + c
f
y
T
(Ax + g(x) x
c
)
(s
c
l
)
T
(x
c
l
c
) (s
c
u
)
T
(u
c
x
c
)
(s
x
l
)
T
(x l
x
) (s
x
u
)
T
(u
x
x).
and the dual problem is given by
maximize L(x
c
, x, y, s
c
l
, s
c
u
, s
x
l
, s
x
u
)
subject to
(x
c
,x)
L(x
c
, x, y, s
c
l
, s
c
u
, s
x
l
, s
x
u
) = 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
which is equivalent to
maximize f(x) y
T
g(x) x
T
(f(x)
T
g(x)
T
y)
+((l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
+ c
f
subject to f(x)
T
+ A
T
y +g(x)
T
y + s
x
l
s
x
u
= c,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
(9.31)
9.5 Recommendations
Often an optimization problem can be formulated in several dierent ways, and the exact formulation
used may have a signicant impact on the solution time and the quality of the solution. In some cases
the dierence between a good and a bad formulation means the ability to solve the problem or
not.
Below is a list of several issues that you should be aware of when developing a good formulation.
1. Sparsity is very important. The constraint matrix A is assumed to be a sparse matrix, where
sparse means that it contains many zeros (typically less than 10% non-zeros). Normally, when
A is sparser, less memory is required to store the problem and it can be solved faster.
2. Avoid large bounds as these can introduce all sorts of numerical problems. Assume that a variable
x
j
has the bounds
0.0 x
j
1.0e16.
The number 1.0e16 is large and it is very likely that the constraint x
j
1.0e16 is non-binding
at optimum, and therefore that the bound 1.0e16 will not cause problems. Unfortunately, this is
a nave assumption because the bound 1.0e16 may actually aect the presolve, the scaling, the
computation of the dual objective value, etc. In this case the constraint x
j
0 is likely to be
sucient, i.e. 1.0e16 is just a way of representing innity.
3. Avoid large penalty terms in the objective, i.e. do not have large terms in the linear part of the
objective function. They will most likely cause numerical problems.
4. On a computer all computations are performed in nite precision, which implies that
1 = 1 +
where is about 10
16
. This means that the results of all computations are truncated and
therefore causing rounding errors. The upshot is that very small numbers and very large numbers
should be avoided, e.g. it is recommended that all elements in A either are zero or belong to the
interval [10
6
, 10
6
]. The same holds for the bounds and the linear objective.
5. Decreasing the number of variables or constraints does not necessarily make it easier to solve
a problem. In certain cases, i.e. in nonlinear optimization, it may be a good idea to introduce
more constraints and variables if it makes the model separable. Furthermore, a big but sparse
problem may be advantageous compared to a smaller but denser problem.
6. Try to avoid linearly dependent rows among the linear constraints. Network ow problems
and multi-commodity network ow problems, for example, often contain one or more linearly
dependent rows.
7. Finally, it is recommended to consult some of the papers about preprocessing to get some ideas
about ecient formulations. See e.g. [3, 4, 14, 15].
9.5.1 Avoid near infeasible models
Consider the linear optimization problem
minimize
subject to x + y 10
10
+ ,
1.0e4x + 2.0e4y 10
6
,
x, y 0.
(9.32)
Clearly, the problem is feasible for = 0. However, for = 1.0e 10 the problem is infeasible.
This implies that an insignicant change in the right side of the constraints makes the problem status
switch from feasible to infeasible. Such a model should be avoided.
9.6 Examples continued
9.6.1 The absolute value
Assume that we have a constraint for the form
[f
T
x + g[ b (9.33)
where x R
n
is a vector of variables, and f R
n
and g, b R are constants.
It is easy to verify that the constraint (9.33) is equivalent to
b f
T
x + g t b (9.34)
which is a set of ordinary linear inequality constraints.
9.6. EXAMPLES CONTINUED 51
Please note that equalities involving an absolute value such as
[x[ = 1
cannot be formulated as a linear or even a as convex nonlinear optimization problem. It requires
integer constraints.
9.6.2 The Markowitz portfolio model
In this section we will show how to model several versions of the Markowitz portfolio model using conic
optimization.
The Markowitz portfolio model deals with the problem of selecting a portfolio of assets, i.e. stocks,
bonds, etc. The goal is to nd a portfolio such that for a given return the risk is minimized. The
assumptions are:
A portfolio can consist of n traded assets numbered 1, 2, . . . held over a period of time.
w
0
j
is the initial holding of asset j where

j
w
0
j
> 0.
r
j
is the return on asset j and is assumed to be a random variable. r has a known mean r and
covariance .
The variable x
j
denotes the amount of asset j traded in the given period of time and has the following
meaning:
If x
j
> 0, then the amount of asset j is increased (by purchasing).
If x
j
< 0, then the amount of asset j is decreased (by selling).
The model deals with two central quantities:
Expected return:
E[r
T
(w
0
+ x)] = r
T
(w
0
+ x).
Variance (Risk):
V [r
T
(w
0
+ x)] = (w
0
+ x)
T
(w
0
+ x).
By denition is positive semi-denite and
Std. dev. =
_
_
_
1
2
(w
0
+ x)
_
_
_
=
_
_
L
T
(w
0
+ x)
_
_
where L is any matrix such that
= LL
T
A low rank of is advantageous from a computational point of view. A valid L can always be computed
as the Cholesky factorization of .
9.6.2.1 Minimizing variance for a given return
In our rst model we want to minimize the variance while selecting a portfolio with a specied expected
target return t. Additionally, the portfolio must satisfy the budget (self-nancing) constraint asserting
that the total amount of assets sold must equal the total amount of assets purchased. This is expressed
in the model
minimize V [r
T
(w
0
+ x)]
subject to E[r
T
(w
0
+ x)] = t,
e
T
x = 0,
(9.35)
where e := (1, . . . , 1)
T
. Using the denitions above this may be formulated as a quadratic optimization
problem:
minimize (w
0
+ x)
T
(w
0
+ x)
subject to r
T
(w
0
+ x) = t,
e
T
x = 0.
(9.36)
9.6.2.2 Conic quadratic reformulation
An equivalent conic quadratic reformulation is given by:
minimize f
subject to
1
2
(w
0
+ x) g = 0,
r
T
(w
0
+ x) = t,
e
T
x = 0,
f |g| .
(9.37)
Here we minimize the standard deviation instead of the variance. Please note that
1
2
can be replaced
by any matrix L where = LL
T
. A low rank L is computationally advantageous.
9.6.2.3 Transaction costs with market impact term
We will now expand our model to include transaction costs as a fraction of the traded volume. [1, pp.
445-475] argues that transaction costs can be modeled as follows
commission +
bid
ask
spread +
trade volume
daily volume
, (9.38)
and that it is important to incorporate these into the model.
In the following we deal with the last of these terms denoted the market impact term. If you sell
(buy) a lot of assets the price is likely to go down (up). This can be captured in the market impact
term
trade volume
daily volume
m
j
_
[x
j
[.
The and daily volume have to be estimated in some way, i.e.
m
j
=

daily volume
9.6. EXAMPLES CONTINUED 53
has to be estimated. The market impact term gives the cost as a fraction of daily traded volume ([x
j
[).
Therefore, the total cost when trading an amount x
j
of asset j is given by
[x
j
[(m
j
[x
j
[
1
2
).
This leads us to the model:
minimize f
subject to
1
2
(w
0
+ x) g = 0,
r
T
(w
0
+ x) = t,
e
T
x + e
T
y = 0,
[x
j
[(m
j
[x
j
[
1
2
) y
j
,
f |g| .
(9.39)
Now, dening the variable transformation
y
j
= m
j
y
j
we obtain
minimize f
subject to
1
2
(w
0
+ x) g = 0,
r
T
(w
0
+ x) = t,
e
T
x + m
T
y = 0,
[x
j
[
3/2
y
j
,
f |g| .
(9.40)
As shown in Section 9.3.3.3 the set
[x
j
[
3/2
y
j
can be modeled by
x
j
z
j
,
x
j
z
j
,
z
2
j
2s
j
y
j
,
u
2
j
2v
j
q
j
,
z
j
= v
j
,
s
j
= u
j
,
q
j
=
1
8
,
q
j
, s
j
, y
j
, v
j
, q
j
0.
(9.41)
9.6.2.4 Further reading
For further reading please see [17] in particular, and [20] and [1], which also contain relevant material.
Chapter 10
The optimizers for continuous
problems
The most essential part of MOSEK is the optimizers. Each optimizer is designed to solve a particular
class of problems i.e. linear, conic, or general nonlinear problems. The purpose of the present chapter
is to discuss which optimizers are available for the continuous problem classes and how the performance
of an optimizer can be tuned, if needed.
This chapter deals with the optimizers for continuous problems with no integer variables.
10.1 How an optimizer works
When the optimizer is called, it roughly performs the following steps:
Presolve: Preprocessing to reduce the size of the problem.
Dualizer: Choosing whether to solve the primal or the dual form of the problem.
Scaling: Scaling the problem for better numerical stability.
Optimize: Solve the problem using selected method.
The rst three preprocessing steps are transparent to the user, but useful to know about for tuning
purposes. In general, the purpose of the preprocessing steps is to make the actual optimization more
ecient and robust.
10.1.1 Presolve
Before an optimizer actually performs the optimization the problem is preprocessed using the so-called
presolve. The purpose of the presolve is to
remove redundant constraints,
eliminate xed variables,
remove linear dependencies,
55
56 CHAPTER 10. THE OPTIMIZERS FOR CONTINUOUS PROBLEMS
substitute out free variables, and
reduce the size of the optimization problem in general.
After the presolved problem has been optimized the solution is automatically postsolved so that the
returned solution is valid for the original problem. Hence, the presolve is completely transparent. For
further details about the presolve phase, please see [3, 4].
It is possible to ne-tune the behavior of the presolve or to turn it o entirely. If presolve consumes
too much time or memory compared to the reduction in problem size gained it may be disabled. This
is done by setting the parameter MSK IPAR PRESOLVE USE to MSK PRESOLVE MODE OFF.
The two most time-consuming steps of the presolve are
the eliminator, and
the linear dependency check.
Therefore, in some cases it is worthwhile to disable one or both of these.
10.1.1.1 Eliminator
The purpose of the eliminator is to eliminate free and implied free variables from the problem using
substitution. For instance, given the constraints
y =

j
x
j
,
y, x 0,
y is an implied free variable that can be substituted out of the problem, if deemed worthwhile.
If the eliminator consumes too much time or memory compared to the reduction in problem size
gained it may be disabled. This can be done with the parameter MSK IPAR PRESOLVE ELIMINATOR USE
to MSK OFF.
10.1.1.2 Linear dependency checker
The purpose of the linear dependency check is to remove linear dependencies among the linear equal-
ities. For instance, the three linear equalities
x
1
+ x
2
+ x
3
= 1,
x
1
+ 0.5x
2
= 0.5,
0.5x
2
+ x
3
= 0.5
contain exactly one linear dependency. This implies that one of the constraints can be dropped without
changing the set of feasible solutions. Removing linear dependencies is in general a good idea since it
reduces the size of the problem. Moreover, the linear dependencies are likely to introduce numerical
problems in the optimization phase.
It is best practise to build models without linear dependencies. If the linear dependencies are
removed at the modeling stage, the linear dependency check can safely be disabled by setting the
parameter MSK IPAR PRESOLVE LINDEP USE to MSK OFF.
10.1. HOW AN OPTIMIZER WORKS 57
10.1.2 Dualizer
All linear, conic, and convex optimization problems have an equivalent dual problem associated with
them. MOSEK has built-in heuristics to determine if it is most ecient to solve the primal or dual
problem. The form (primal or dual) solved is displayed in the MOSEK log. Should the internal
heuristics not choose the most ecient form of the problem it may be worthwhile to set the dualizer
manually by setting the parameters:
MSK IPAR INTPNT SOLVE FORM: In case of the interior-point optimizer.
MSK IPAR SIM SOLVE FORM: In case of the simplex optimizer.
Note that currently only linear problems may be dualized.
10.1.3 Scaling
Problems containing data with large and/or small coecients, say 1.0e +9 or 1.0e 7, are often hard
to solve. Signicant digits may be truncated in calculations with nite precision, which can result in
the optimizer relying on inaccurate calculations. Since computers work in nite precision, extreme
coecients should be avoided. In general, data around the same order of magnitude is preferred,
and we will refer to a problem, satisfying this loose property, as being well-scaled. If the problem is
not well scaled, MOSEK will try to scale (multiply) constraints and variables by suitable constants.
MOSEK solves the scaled problem to improve the numerical properties.
The scaling process is transparent, i.e. the solution to the original problem is reported. It is
important to be aware that the optimizer terminates when the termination criterion is met on the
scaled problem, therefore signicant primal or dual infeasibilities may occur after unscaling for badly
scaled problems. The best solution to this problem is to reformulate it, making it better scaled.
By default MOSEK heuristically chooses a suitable scaling. The scaling for interior-point and
simplex optimizers can be controlled with the parameters
MSK IPAR INTPNT SCALING and MSK IPAR SIM SCALING
respectively.
10.1.4 Using multiple CPUs
The interior-point optimizers in MOSEK have been parallelized. This means that if you solve linear,
quadratic, conic, or general convex optimization problem using the interior-point optimizer, you can
take advantage of multiple CPUs.
By default MOSEK uses one thread to solve the problem, but the number of threads (and thereby
CPUs) employed can be changed by setting the parameter MSK IPAR INTPNT NUM THREADS This should
never exceed the number of CPUs on the machine.
The speed-up obtained when using multiple CPUs is highly problem and hardware dependent, and
consequently, it is advisable to compare single threaded and multi threaded performance for the given
problem type to determine the optimal settings.
For small problems, using multiple threads will probably not be worthwhile.
10.2 Linear optimization
10.2.1 Optimizer selection
Two dierent types of optimizers are available for linear problems: The default is an interior-point
method, and the alternatives are simplex methods. The optimizer can be selected using the parameter
MSK IPAR OPTIMIZER.
10.2.2 The interior-point optimizer
The purpose of this section is to provide information about the algorithm employed in MOSEK interior-
point optimizer.
In order to keep the discussion simple it is assumed that MOSEK solves linear optimization prob-
lems on standard form
minimize c
T
x
subject to Ax = b,
x 0.
(10.1)
This is in fact what happens inside MOSEK; for eciency reasons MOSEK converts the problem to
standard form before solving, then convert it back to the input form when reporting the solution.
Since it is not known beforehand whether problem (10.1) has an optimal solution, is primal infeasible
or is dual infeasible, the optimization algorithm must deal with all three situations. This is the reason
that MOSEK solves the so-called homogeneous model
Ax b = 0,
A
T
y + s c = 0,
c
T
x + b
T
y = 0,
x, s, , 0,
(10.2)
where y and s correspond to the dual variables in (10.1), and and are two additional scalar variables.
Note that the homogeneous model (10.2) always has solution since
(x, y, s, , ) = (0, 0, 0, 0, 0)
is a solution, although not a very interesting one.
Any solution
(x
, y
, s
)
to the homogeneous model (10.2) satises
x
j
s
j
= 0 and
= 0.
Moreover, there is always a solution that has the property
> 0.
First, assume that
> 0. It follows that

A
x
= b,
A
T y
+
s
= c,
c
T x
+ b
T y
= 0,
x
, s
0.
(10.3)
This shows that
x
is a primal optimal solution and (

y
,
s
) is a dual optimal solution; this is reported

as the optimal interior-point solution since
(x, y, s) =
_
x
,
y
,
s
_
is a primal-dual optimal solution.
On other hand, if
> 0 then
Ax
= 0,
A
T
y
+ s
= 0,
c
T
x
+ b
T
y
,
x
, s
0.
(10.4)
This implies that at least one of
c
T
x
> 0 (10.5)
or
b
T
y
> 0 (10.6)
is satised. If (10.5) is satised then x
is a certicate of dual infeasibility, whereas if (10.6) is satised

then y
is a certicate of dual infeasibility.

In summary, by computing an appropriate solution to the homogeneous model, all information
required for a solution to the original problem is obtained. A solution to the homogeneous model can
be computed using a primal-dual interior-point algorithm [9].
10.2.2.1 Interior-point termination criterion
For eciency reasons it is not practical to solve the homogeneous model exactly. Hence, an exact
optimal solution or an exact infeasibility certicate cannot be computed and a reasonable termination
criterion has to be employed.
In every iteration, k, of the interior-point algorithm a trial solution
(x
k
, y
k
, s
k
,
k
,
k
)
to homogeneous model is generated where
x
k
, s
k
,
k
,
k
> 0.
Whenever the trial solution satises the criterion
_
_
_A
x
k
k
b
_
_
_
p
(1 +|b|),
_
_
_A
T x
k
k
+
s
k
k
c
_
_
_
d
(1 +|c|), and
min
_
(x
k
)
T
s
k
+
k
k
(
k
)
2
,
c
T
x
k
k

b
T
y
k
_

g
max
_
1,
c
T
x
k
_
,
(10.7)
the interior-point optimizer is terminated and
(x
k
, y
k
, s
k
)
k
is reported as the primal-dual optimal solution. The interpretation of (10.7) is that the optimizer is
terminated if
Tolerance Parameter name
p
MSK DPAR INTPNT TOL PFEAS
d
MSK DPAR INTPNT TOL DFEAS
g
MSK DPAR INTPNT TOL REL GAP
i
MSK DPAR INTPNT TOL INFEAS
Table 10.1: Parameters employed in termination criterion.
x
k
k
is approximately primal feasible,
_
y
k
k
,
s
k
k
_
is approximately dual feasible, and
the duality gap is almost zero.
On the other hand, if the trial solution satises
i
c
T
x
k
>
|c|
max(|b| , 1)
_
_
Ax
k
_
_
(10.8)
then the problem is declared dual infeasible and x
k
is reported as a certicate of dual infeasibility.
The motivation for this stopping criterion is as follows: First assume that
_
_
Ax
k
_
_
= 0; then x
k
is an
exact certicate of dual infeasibility. Next assume that this is not the case, i.e.
_
_
Ax
k
_
_
> 0,
and dene
x :=
i
max(1, |b|)x
k
|Ax
k
| |c|
.
It is easy to verify that
|A x| =
i
and c
T
x > 1,
which shows x is an approximate certicate dual infeasibility where
i
controls the quality of the
approximation. A smaller value means a better approximation.
Finally, if
i
b
T
y
k
|b|
max(1, |c|)
_
_
A
T
y
k
+ s
k
_
_
(10.9)
then y
k
is reported as a certicate of primal infeasibility.
It is possible to adjust the tolerances
p
,
d
,
g
and
i
using parameters; see table 10.1 for details.
The default values of the termination tolerances are chosen such that for a majority of problems
appearing in practice it is not possible to achieve much better accuracy. Therefore, tightening the
tolerances usually is not worthwhile. However, an inspection of (10.7) reveals that quality of the
solution is dependent on |b| and |c|; the smaller the norms are, the better the solution accuracy.
The interior-point method as implemented by MOSEK will converge toward optimality and primal
and dual feasibility at the same rate [9]. This means that if the optimizer is stopped prematurely then
it is very unlikely that either the primal or dual solution is feasible. Another consequence is that in
most cases all the tolerances,
p
,
d
and
g
, has to be relaxed together to achieve an eect.
The basis identication discussed in section 10.2.2.2 requires an optimal solution to work well;
hence basis identication should turned o if the termination criterion is relaxed.
To conclude the discussion in this section, relaxing the termination criterion is usually is not
worthwhile.
10.2.2.2 Basis identication
An interior-point optimizer does not return an optimal basic solution unless the problem has a
unique primal and dual optimal solution. Therefore, the interior-point optimizer has an optional
post-processing step that computes an optimal basic solution starting from the optimal interior-point
solution. More information about the basis identication procedure may be found in [6].
Please note that a basic solution is often more accurate than an interior-point solution.
By default MOSEK performs a basis identication. However, if a basic solution is not needed, the
basis identication procedure can be turned o. The parameters
MSK IPAR INTPNT BASIS,
MSK IPAR BI IGNORE MAX ITER, and
MSK IPAR BI IGNORE NUM ERROR
controls when basis identication is performed.
10.2.2.3 The interior-point log
Below is a typical log output from the interior-point optimizer presented:
Optimizer - threads : 1
Optimizer - solved problem : the dual
Factor - setup time : 0.04 order time : 0.00
Factor - GP order used : no GP order time : 0.00
Factor - nonzeros before factor : 3 after factor : 3
Factor - offending columns : 0 flops : 1.70e+001
ITE PFEAS DFEAS KAP/TAU POBJ DOBJ MU TIME
0 2.0e+002 2.9e+001 2.0e+002 -0.000000000e+000 -1.204741644e+003 2.0e+002 0.44
1 2.2e+001 3.1e+000 7.3e+002 -5.885951891e+003 -5.856764353e+003 2.2e+001 0.57
2 3.8e+000 5.4e-001 9.7e+001 -7.405187479e+003 -7.413054916e+003 3.8e+000 0.58
3 4.0e-002 5.7e-003 2.6e-001 -7.664507945e+003 -7.665313396e+003 4.0e-002 0.58
4 4.2e-006 6.0e-007 2.7e-005 -7.667999629e+003 -7.667999714e+003 4.2e-006 0.59
5 4.2e-010 6.0e-011 2.7e-009 -7.667999994e+003 -7.667999994e+003 4.2e-010 0.59
The rst line displays the number of threads used by the optimizer and second line tells that the
optimizer choose to solve the dual problem rather the primal problem. The next line displays the
problem dimensions as seen by the optimizer, and the Factor... lines show various statistics. This
is followed by the iteration log.
Using the same notation as in section 10.2.2 the columns of the iteration log has the following
meaning:
ITE: Iteration index.
PFEAS:
_
_
Ax
k
b
k
_
_
. The numbers in this column should converge monotonically towards to
zero.
DFEAS:
_
_
A
T
y
k
+ s
k
c
k
_
_
. The numbers in this column should converge monotonically toward
to zero.
KAP/TAU:
k
/
k
. If the numbers in this column converge toward zero then the problem has an
optimal solution. Otherwise if the numbers converge towards innity, the problem is primal
or/and dual infeasible.
POBJ: c
T
x
k
/
k
. An estimate for the primal objective value.
DOBJ: b
T
y
k
/
k
. An estimate for the dual objective value.
MU:
(x
k
)
T
s
k
+
k
k
n+1
. The numbers in this column should always converge monotonically to zero.
TIME: Time spend since the optimization started.
10.2.3 The simplex based optimizer
An alternative to the interior-point optimizer is the simplex optimizer.
The simplex optimizer uses a dierent method that allows exploiting an initial guess for the optimal
solution to reduce the solution time. Depending on the problem it may be faster or slower to use an
initial guess; see section 10.2.4 for a discussion.
MOSEK provides both a primal and a dual variant of the simplex optimizer we will return to
this later.
10.2.3.1 Simplex termination criterion
The simplex optimizer terminates when it nds an optimal basic solution or an infeasibility certicate.
A basic solution is optimal when it is primal and dual feasible; see (9.1) and (9.2) for a denition
of the primal and dual problem. Due the fact that to computations are performed in nite pre-
cision MOSEK allows violation of primal and dual feasibility within certain tolerances. The user
can control the allowed primal and dual infeasibility with the parameters MSK DPAR BASIS TOL X and
MSK DPAR BASIS TOL S.
10.2.3.2 Starting from an existing solution
When using the simplex optimizer it may be possible to reuse an existing solution and thereby reduce
the solution time signicantly. When a simplex optimizer starts from an existing solution it is said to
perform a hot-start. If the user is solving a sequence of optimization problems by solving the problem,
making modications, and solving again, MOSEK will hot-start automatically.
Setting the parameter MSK IPAR OPTIMIZER to MSK OPTIMIZER FREE SIMPLEX instructs MOSEK to
select automatically between the primal and the dual simplex optimizers. Hence, MOSEK tries to
choose the best optimizer for the given problem and the available solution.
By default MOSEK uses presolve when performing a hot-start. If the optimizer only needs very
few iterations to nd the optimal solution it may be better to turn o the presolve.
10.2.3.3 Numerical diculties in the simplex optimizers
Though MOSEK is designed to minimize numerical instability, completely avoiding it is impossible
when working in nite precision. MOSEK counts a numerical unexpected behavior event inside the
optimizer as a set-back. The user can dene how many set-backs the optimizer accepts; if that number
is exceeded, the optimization will be aborted. Set-backs are implemented to avoid long sequences
where the optimizer tries to recover from an unstable situation.
Set-backs are, for example, repeated singularities when factorizing the basis matrix, repeated loss
of feasibility, degeneracy problems (no progress in objective) and other events indicating numerical
diculties. If the simplex optimizer encounters a lot of set-backs the problem is usually badly scaled;
in such a situation try to reformulate into a better scaled problem. Then, if a lot of set-backs still
occur, trying one or more of the following suggestions may be worthwhile:
Raise tolerances for allowed primal or dual feasibility: Hence, increase the value of
MSK DPAR BASIS TOL X, and
MSK DPAR BASIS TOL S.
Raise or lower pivot tolerance: Change the MSK DPAR SIMPLEX ABS TOL PIV parameter.
Switch optimizer: Try another optimizer.
Switch o crash: Set both MSK IPAR SIM PRIMAL CRASH and MSK IPAR SIM DUAL CRASH to 0.
Experiment with other pricing strategies: Try dierent values for the parameters
MSK IPAR SIM PRIMAL SELECTION and
MSK IPAR SIM DUAL SELECTION.
If you are using hot-starts, in rare cases switching o this feature may improve stability. This is
controlled by the MSK IPAR SIM HOTSTART parameter.
Increase maximum set-backs allowed controlled by MSK IPAR SIM MAX NUM SETBACKS.
If the problem repeatedly becomes infeasible try switching o the special degeneracy handling.
See the parameter MSK IPAR SIM DEGEN for details.
10.2.4 The interior-point or the simplex optimizer?
Given a linear optimization problem, which optimizer is the best: The primal simplex, the dual simplex
or the interior-point optimizer?
It is impossible to provide a general answer to this question, however, the interior-point optimizer
behaves more predictably it tends to use between 20 and 100 iterations, almost independently of
problem size but cannot perform hot-start, while simplex can take advantage of an initial solution,
but is less predictable for cold-start. The interior-point optimizer is used by default.
10.2.5 The primal or the dual simplex variant?
MOSEK provides both a primal and a dual simplex optimizer. Predicting which simplex optimizer
is faster is impossible, however, in recent years the dual optimizer has seen several algorithmic and
computational improvements, which, in our experience, makes it faster on average than the primal
simplex optimizer. Still, it depends much on the problem structure and size.
Setting the MSK IPAR OPTIMIZER parameter to MSK OPTIMIZER FREE SIMPLEX instructs MOSEK to
choose which simplex optimizer to use automatically.
To summarize, if you want to know which optimizer is faster for a given problem type, you should
try all the optimizers.
10.3 Linear network optimization
10.3.1 Network ow problems
MOSEK includes a network simplex solver which, on avarage, solves network problems 10 to 100 times
faster than the standard simplex optimizers.
To use the network simplex optimizer, do the following:
Input the network ow problem as an ordinary linear optimization problem.
Set the parameters
MSK IPAR SIM NETWORK DETECT to 0, and
MSK IPAR OPTIMIZER to MSK OPTIMIZER FREE SIMPLEX.
MOSEK will automatically detect the network structure and apply the specialized simplex optimizer.
10.3.2 Embedded network problems
Often problems contains both large parts with network structure and some non-network constraints
or variables such problems are said to have embedded network structure.
If the procedure described in section 10.3.1 is applied, MOSEK will attemt to exploit this structure
to speed up the optimization.
This is done heuristically by detecting the largest network embedded in the problem, solving this
subproblem using the network simplex optimizer, and using the solution to hot-start a normal simplex
optimizer.
The MSK IPAR SIM NETWORK DETECT parameter denes how large a percentage of the problem should
be a network before the specialized solver is applied. In general, it is recommended to use the network
optimizer only on problems containing a substantial embedded network.
If MOSEK only nds limited network structure in a problem, consider trying to switch o presolve
MSK IPAR PRESOLVE USE and scaling MSK IPAR SIM SCALING, since in rare cases it might disturb the
network heuristic.
Parameter name Purpose
MSK DPAR INTPNT CO TOL PFEAS Controls primal feasibility
MSK DPAR INTPNT CO TOL DFEAS Controls dual feasibility
MSK DPAR INTPNT CO TOL REL GAP Controls relative gap
MSK DPAR INTPNT TOL INFEAS Controls when the problem is declared infeasible
MSK DPAR INTPNT CO TOL MU RED Controls when the complementarity is reduced enough
Table 10.2: Parameters employed in termination criterion.
10.4 Conic optimization
For conic optimization problems only an interior-point type optimizer is available. The interior-point
optimizer is an implementation of the so-called homogeneous and self-dual algorithm. For a detailed
description of the algorithm, please see [5].
10.4.1.1 Interior-point termination criteria
The parameters controlling when the conic interior-point optimizer terminates are shown in Table 10.2.
10.5 Nonlinear convex optimization
For quadratic, quadratically constrained, and general convex optimization problems an interior-point
type optimizer is available. The interior-point optimizer is an implementation of the homogeneous and
self-dual algorithm. For a detailed description of the algorithm, please see [7, 8].
10.5.1.1 The convexity requirement
Continuous nonlinear problems are required to be convex. For quadratic problems MOSEK test this
requirement before optimizing. Specifying a non-convex problem results in an error message.
The following parameters are available to control the convexity check:
MSK IPAR CHECK CONVEXITY: Turn convexity check on/o.
MSK DPAR CHECK CONVEXITY REL TOL: Tolerance for convexity check.
MSK IPAR LOG CHECK CONVEXITY: Turn on more log information for debugging.
10.5.1.2 The dierentiabilty requirement
The nonlinear optimizer in MOSEK requires both rst order and second order derivatives. This of
course implies care should be taken when solving problems involving non-dierentiable functions.
For instance, the function
f(x) = x
2
Parameter name Purpose
MSK DPAR INTPNT NL TOL PFEAS Controls primal feasibility
MSK DPAR INTPNT NL TOL DFEAS Controls dual feasibility
MSK DPAR INTPNT NL TOL REL GAP Controls relative gap
MSK DPAR INTPNT TOL INFEAS Controls when the problem is declared infeasible
MSK DPAR INTPNT NL TOL MU RED Controls when the complementarity is reduced enough
Table 10.3: Parameters employed in termination criteria.
is dierentiable everywhere whereas the function
f(x) =

x
is only direntiable for x > 0. In order to make sure that MOSEK evaulates the functions at points
where they are dierentiable, the function domains must be dened by setting appropriate variable
bounds.
In general, if a variable is not ranged MOSEK will only evaluate that variable at points strictly
within the bounds. Hence, imposing the bound
x 0
in the case of

x is sucient to guarantee that the function will only be evaluated in points where it
is dierentiable.
However, if a function is dierentiable on closed a range, specifying the variable bounds is not
sucient. Consider the function
f(x) =
1
x
+
1
1 x
. (10.10)
In this case the bounds
0 x 1
will not guarantee that MOSEK only evalues the function for x between 0 and 1. To force MOSEK to
strictly satisfy both bounds on ranged variables set the parameter MSK IPAR INTPNT STARTING POINT
to MSK STARTING POINT SATISFY BOUNDS.
For eciency reasons it may be better to reformulate the problem than to force MOSEK to observe
ranged bounds strictly. For instance, (10.10) can be reformulated as follows
f(x) =
1
x
+
1
y
0 = 1 x y
0 x
0 y.
10.5.1.3 Interior-point termination criteria
The parameters controlling when the general convex interior-point optimizer terminates are shown in
Table 10.3.
10.6. SOLVING PROBLEMS IN PARALLEL 67
10.6 Solving problems in parallel
If a computer has multiple CPUs, or has a CPU with multiple cores, it is possible for MOSEK to take
advantage of this to speed up solution times.
10.6.1 Thread safety
The MOSEK API is thread-safe provided that a task is only modied or accessed from one thread at
any given time accessing two separate tasks from two separate threads at the same time is safe.
Sharing an environment between threads is safe.
10.6.2 The parallelized interior-point optimizer
The interior-point optimizer is capable of using multiple CPUs or cores. This implies that whenever
the MOSEK interior-point optimizer solves an optimization problem, it will try to divide the work so
that each CPU gets a share of the work. The user decides how many CPUs MOSEK should exploit.
It is not always possible to divide the work equally, and often parts of the computations and the
coordination of the work is processed sequentially, even if several CPUs are present. Therefore, the
speed-up obtained when using multiple CPUs is highly problem dependent. However, as a rule of
thumb, if the problem solves very quickly, i.e. in less than 60 seconds, it is not advantageous to use
the parallel option.
The MSK IPAR INTPNT NUM THREADS parameter sets the number of threads (and therefore the num-
ber of CPUs) that the interior point optimizer will use.
10.6.3 The concurrent optimizer
An alternative to the parallel interior-point optimizer is the concurrent optimizer. The idea of the
concurrent optimizer is to run multiple optimizers on the same problem concurrently, for instance,
it allows you to apply the interior-point and the dual simplex optimizers to a linear optimization
problem concurrently. The concurrent optimizer terminates when the rst of the applied optimizers
has terminated successfully, and it reports the solution of the fastest optimizer. In that way a new
optimizer has been created which essentially performs as the fastest of the interior-point and the
dual simplex optimizers.Hence, the concurrent optimizer is the best one to use if there are multiple
optimizers available in MOSEK for the problem and you cannot say beforehand which one will be
faster.
Note in particular that any solution present in the task will also be used for hot-starting the simplex
algorithms. One possible scenario would therefore be running a hot-start dual simplex in parallel with
interior point, taking advantage of both the stability of the interior-point method and the ability of
the simplex method to use an initial solution.
By setting the
MSK_IPAR_OPTIMIZER
parameter to
MSK_OPTIMIZER_CONCURRENT
the concurrent optimizer chosen.
The number of optimizers used in parallel is determined by the
Optimizer Associated Default
parameter priority
MSK OPTIMIZER INTPNT MSK IPAR CONCURRENT PRIORITY INTPNT 4
MSK OPTIMIZER FREE SIMPLEX MSK IPAR CONCURRENT PRIORITY FREE SIMPLEX 3
MSK OPTIMIZER PRIMAL SIMPLEX MSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEX 2
MSK OPTIMIZER DUAL SIMPLEX MSK IPAR CONCURRENT PRIORITY DUAL SIMPLEX 1
Table 10.4: Default priorities for optimizer selection in concurrent optimization.
MSK_IPAR_CONCURRENT_NUM_OPTIMIZERS.
parameter. Moreover, the optimizers are selected according to a preassigned priority with optimizers
having the highest priority being selected rst. The default priority for each optimizer is shown in
Table 10.6.3. For example, setting the MSK IPAR CONCURRENT NUM OPTIMIZERS parameter to 2 tells the
concurrent optimizer to the apply the two optimizers with highest priorities: In the default case that
means the interior-point optimizer and one of the simplex optimizers.
10.6.3.1 Concurrent optimization from the command line
The command line
mosek afiro.mps -d MSK_IPAR_OPTIMIZER MSK_OPTIMIZER_CONCURRENT \
-d MSK_IPAR_CONCURRENT_NUM_OPTIMIZERS 2
produces the following (edited) output:
...
Number of concurrent optimizers : 2
Optimizer selected for thread number 0 : interior-point (threads = 1)
Optimizer selected for thread number 1 : free simplex
Total number of threads required : 2
...
Thread number 1 (free simplex) terminated first.
...
Concurrent optimizer terminated. CPU Time: 0.03. Real Time: 0.00.
As indicated in the log information, the interior-point and the free simplex optimizers are employed
concurrently. However, only the output from the optimizer having the highest priority is printed to
the screen. In the example this is the interior-point optimizer.
The line
Total number of threads required : 2
10.7. UNDERSTANDING SOLUTION QUALITY 69
indicates the number of threads used. If the concurrent optimizer should be eective, this should be
lower than the number of CPUs.
In the above example the simplex optimizer nishes rst as indicated in the log information.
10.7 Understanding solution quality
MOSEK will, in general, not produce an exact optimal solution; for eciency reasons computations are
performed in nite precision. This means that it is important to evaluate the quality of the reported
solution. To evaluate the solution quality inspect the following properties:
The solution status reported by MOSEK.
Primal feasibility: How much the solution violates the original constraints of the problem.
Dual feasibility: How much the dual solution violates the constraints of the dual problem.
Duality gap: The dierence between the primal and dual objective values.
Ideally, the primal and dual solutions should only violate the constraints of their respective problem
slightly and the primal and dual objective values should be close. This should be evaluated in the
context of the problem: How good is the data precision in the problem, and how exact a solution is
required.
10.7.1 The solution summary
The solution summary is a small display generated by MOSEK that makes it easy to check the quality
of the solution.
10.7.1.1 The optimal case
The solution summary has the format
Primal - objective: 5.5018458883e+03 eq. infeas.: 1.20e-12 max bound infeas.: 2.31e-14
Dual - objective: 5.5018458883e+03 eq. infeas.: 1.15e-14 max bound infeas.: 7.11e-15
i.e. it shows status information, objective values and quality measures for the primal and dual solutions.
Assumeing that we are solving a linear optimization problem and referring to the problems (9.1)
and (9.2), the interpretation of the solution summary is as follows:
Problem status: The status of the problem.
Solution status: The status of the solution.
Primal objective: The primal objective value.
Primal eq. infeas: |Ax
x
x
c
|
.
Primal max bound infeas.: max(l
c
x
c
; x
c
u
c
; l
x
x
x
; x
x
u
x
; 0).
Dual objective: The dual objective value.
Dual eq. infeas:
_
_
y + s
c
l
s
c
u
; A
T
y + s
x
l
s
x
u
c
_
_
.
Dual max bound infeas.: max(s
c
l
; s
c
u
; s
x
l
; s
x
u
; 0).
In the solution summary above the solution is classied as OPTIMAL, meaning that the solution
should be a good approximation to the true optimal solution. This seems very reasonable since the
primal and dual solutions only violate their respective constraints slightly. Moreover, the duality gap
is small, i.e. the primal and dual objective values are almost identical.
10.7.1.2 The primal infeasible case
For an infeasible problem the solution summary might look like this:
Problem status : PRIMAL_INFEASIBLE
Solution status : PRIMAL_INFEASIBLE_CER
Dual - objective: 1.0000000000e+02 eq. infeas.: 0.00e+00 max bound infeas.: 0.00e+00
It is known that if the problem is primal infeasible then an infeasibility certicate exists, which is
a solution to the problem (9.3) having a positive objective value. Note that the primal solution plays
no role and only the dual solution is used to specify the certicate.
Therefore, in the primal infeasible case the solution summery should report how good the dual
solution is to the problem (9.3). The interpretation of the solution summary is as follows:
Problem status: The status of the problem.
Solution status: The status of the solution.
Primal objective: Should be ignored.
Primal eq. infeas: Should be ignored.
Primal max bound infeas.: Should be ignored.
Dual objective: (l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
.
Dual eq. infeas:
_
_
y + s
c
l
s
c
u
; A
T
y + s
x
l
s
x
u
0
_
_
.
Dual max bound infeas.: max(s
c
l
; s
c
u
; s
x
l
; s
x
u
).
Please note that
any information about the primal solution should be ignored.
the dual objective value should be strictly positive if primal problem is minimization problem.
Otherwise it should be strictly negative.
the bigger the ratio
(l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
max(|y + s
c
l
s
c
u
; A
T
y + s
x
l
s
x
u
0|
, max(s
c
l
; s
c
u
; s
x
l
; s
x
u
))
is, the better the certicate is. The reason is that a certicate is a ray, and hence only the
direction is important. Therefore, in principle, the certicate should be normalized before using
it.
Please see Section 12.2 for more information about certicates of infeasibility.
Chapter 11
The optimizer for mixed integer
problems
A problem is a mixed-integer optimization problem when one or more of the variables are constrained
to be integers. The integer optimizer available in MOSEK can solve integer optimization problems
involving
linear,
quadratic and
conic
constraints. However, a problem is not allowed to have both conic constraints and quadratic objective
or constraints.
Readers unfamiliar with integer optimization are strongly recommended to consult some relevant
literature, e.g. the book [23] by Wolsey is a good introduction to integer optimization.
11.1 Some notation
In general, an integer optimization problem has the form
z
= minimize c
T
x
subject to l
c
Ax u
c
,
l
x
Ax u
x
,
x
j
Z, j ,
(11.1)
where is an index set specifying which variables are integer-constrained. Frequently we talk about
the continuous relaxation of an integer optimization problem dened as
z = minimize c
T
x
subject to l
c
Ax u
c
,
l
x
Ax u
x
(11.2)
73
74 CHAPTER 11. THE OPTIMIZER FOR MIXED INTEGER PROBLEMS
i.e. we ignore the constraint
x
j
Z, j .
Moreover, let x be any feasible solution to (11.1) and dene
z := c
T
x.
It should be obvious that
z z
z
holds. This is an important observation since if we assume that it is not possible to solve the mixed-
integer optimization problem within a reasonable time frame, but that a feasible solution can be found,
then the natural question is: How far is the obtained solution from the optimal solution? The answer
is that no feasible solution can have an objective value smaller than z, which implies that the obtained
solution is no further away from the optimum than z z.
11.2 An important fact about integer optimization problems
It is important to understand that in a worst-case scenario, the time required to solve integer optimiza-
tion problems grows exponentially with the size of the problem. For instance, assume that a problem
contains n binary variables, then the time required to solve the problem in the worst case may be
proportional to 2
n
. It is a simple exercise to verify that 2
n
is huge even for moderate values of n.
In practice this implies that the focus should be on computing a near optimal solution quickly
rather than at locating an optimal solution.
11.3 How the integer optimizer works
The process of solving an integer optimization problem can be split in three phases:
Presolve: In this phase the optimizer tries to reduce the size of the problem using preprocessing
techniques. Moreover, it strengthens the continuous relaxation, if possible.
Heuristic: Using heuristics the optimizer tries to guess a good feasible solution.
Optimization: The optimal solution is located using a variant of the branch-and-cut method.
In some cases the integer optimizer may locate an optimal solution in the preprocessing stage or
conclude that the problem is infeasible. Therefore, the heuristic and optimization stages may never be
performed.
11.3.1 Presolve
In the preprocessing stage redundant variables and constraints are removed. The presolve stage can
be turned o using the MSK IPAR MIO PRESOLVE USE parameter.
11.4. TERMINATION CRITERION 75
11.3.2 Heuristic
Initially, the integer optimizer tries to guess a good feasible solution using dierent heuristics:
First a very simple rounding heuristic is employed.
Next, if deemed worthwhile, the feasibility pump heuristic is used.
Finally, if the two previous stages did not produce a good initial solution, more sophisticated
heuristics are used.
The following parameters can be used to control the eort made by the integer optimizer to nd
an initial feasible solution.
MSK IPAR MIO HEURISTIC LEVEL: Controls how sophisticated and computationally expensive a
heuristic to employ.
MSK DPAR MIO HEURISTIC TIME: The minimum amount of time to spend in the heuristic search.
MSK IPAR MIO FEASPUMP LEVEL: Controls how aggressively the feasibility pump heuristic is used.
11.3.3 The optimization phase
This phase solves the problem using the branch and cut algorithm.
11.4 Termination criterion
In general, it is impossible to nd an exact feasible and optimal solution to an integer optimization
problem in a reasonable amount of time, though in many practical cases it may be possible. There-
fore, the integer optimizer employs a relaxed feasibility and optimality criterion to determine when a
satisfactory solution is located.
A candidate solution, i.e. a solution to (11.2), is said to be an integer feasible solution if the
criterion
min([x
j
[ x
j
|, x
j
| [x
j
[) max(
1
,
2
[x
j
[) j
is satised. Hence, such a solution is dened as a feasible solution to (11.1).
Whenever the integer optimizer locates an integer feasible solution it will check if the criterion
z z max(
3
,
4
max(1, [z[))
is satised. If this is the case, the integer optimizer terminates and reports the integer feasible solution
as an optimal solution. Please note that z is a valid lower bound determined by the integer optimizer
during the solution process, i.e.
z z
.
The lower bound z normally increases during the solution process.
The tolerances can are specied using parameters see Table 11.1. If an optimal solution cannot
be located within a reasonable time, it may be advantageous to employ a relaxed termination criterion
after some time. Whenever the integer optimizer locates an integer feasible solution and has spent at
Tolerance Parameter name
1
MSK DPAR MIO TOL ABS RELAX INT
2
MSK DPAR MIO TOL REL RELAX INT
3
MSK DPAR MIO TOL ABS GAP
4
MSK DPAR MIO TOL REL GAP
5
MSK DPAR MIO NEAR TOL ABS GAP
6
MSK DPAR MIO NEAR TOL REL GAP
Table 11.1: Integer optimizer tolerances.
Parameter name Delayed Explanation
MSK IPAR MIO MAX NUM BRANCHES Yes Maximum number of branches allowed.
MSK IPAR MIO MAX NUM RELAXS Yes Maximum number of relaxations allowed.
MSK IPAR MIO MAX NUM SOLUTIONS Yes Maximum number of feasible integer solutions allowed.
Table 11.2: Parameters aecting the termination of the integer optimizer.
least the number of seconds dened by the MSK DPAR MIO DISABLE TERM TIME parameter on solving
the problem, it will check whether the criterion
z z max(
5
,
6
max(1, [z[))
is satised. If it is satised, the optimizer will report that the candidate solution is near optimal and
then terminate. All tolerances can be adjusted using suitable parameters see Table 11.1. In Table
11.2 some other parameters aecting the integer optimizer termination criterion are shown. Please
note that if the eect of a parameter is delayed, the associated termination criterion is applied only
after some time, specied by the MSK DPAR MIO DISABLE TERM TIME parameter.
11.5 How to speed up the solution process
As mentioned previously, in many cases it is not possible to nd an optimal solution to an integer
optimization problem in a reasonable amount of time. Some suggestions to reduce the solution time
are:
Relax the termination criterion: In case the run time is not acceptable, the rst thing to do is
to relax the termination criterion see Section 11.4 for details.
Specify a good initial solution: In many cases a good feasible solution is either known or easily
computed using problem specic knowledge. If a good feasible solution is known, it is usually
worthwhile to use this as a starting point for the integer optimizer.
Improve the formulation: A mixed-integer optimization problem may be impossible to solve
in one form and quite easy in another form. However, it is beyond the scope of this manual
to discuss good formulations for mixed-integer problems. For discussions on this topic see for
example [23].
11.6 Understanding solution quality
To determine the quality of the solution one should check the following:
The solution status key returned by MOSEK.
The optimality gap: A messure for how much the located solution can deviate from the optimal
solution to the problem.
Feasibility. How much the solution violates the constraints of the problem.
The optimality gap is a measure for how close the solution is to the optimal solution. The optimality
gap is given by
= [(objective value of feasible solution) (objective bound)[.
The objective value of the solution is guarentted to be within of the optimal solution.
The optimality gap can be retrived through the solution item MSK DINF MIO OBJ ABS GAP. Often it
is more meaningful to look at the optimality gap normalized with the magnitude of the solution. The
relative optimality gap is available in MSK DINF MIO OBJ REL GAP.
11.6.1 Solutionsummary
After a call to the optimizer the solution summary might look like this:
Problem status : PRIMAL_FEASIBLE
Solution status : INTEGER_OPTIMAL
cone infeas.: 0.00e+00 integer infeas.: 0.00e+00
The second line contains the solution status key. This shows how MOSEK classied the solution.
In this case it is INTEGER OPTIMAL, meaning that the solution is considered to be optimal within the
selected tolerances.
The third line contains information relating to the solution. The rst number is the primal objective
function. The second and third number is the maximum infeasibility in the equality constraints and
bounds respectfully. The fourth and fth number is the maximum infeasibility in the conic and integral
contraints. All the numbers relating to the feasibility of the solution should be small for the solution
to be valid.
Chapter 12
The analyzers
12.1 The problem analyzer
The problem analyzer prints a detailed survey of the models
linear constraints and objective
quadratic constraints
conic constraints
variables
In the initial stages of model formulation the problem analyzer may be used as a quick way of
verifying that the model has been built or imported correctly. In later stages it can help revealing
special structures within the model that may be used to tune the optimizers performance or to identify
the causes of numerical diculties.
The problem analyzer is run from the command line using the -anapro argument and produces
something similar to the following (this is the problemanalyzers survey of the aflow30a problem from
the MIPLIB 2003 collection, see Appendix J for more examples):
Analyzing the problem
Constraints Bounds Variables
upper bd: 421 ranged : all cont: 421
fixed : 58 bin : 421
-------------------------------------------------------------------------------
Objective, min cx
range: min |c|: 0.00000 min |c|>0: 11.0000 max |c|: 500.000
distrib: |c| vars
0 421
[11, 100) 150
[100, 500] 271
-------------------------------------------------------------------------------
79
80 CHAPTER 12. THE ANALYZERS
Constraint matrix A has
479 rows (constraints)
842 columns (variables)
2091 (0.518449%) nonzero entries (coefficients)
Row nonzeros, A_i
range: min A_i: 2 (0.23753%) max A_i: 34 (4.038%)
distrib: A_i rows rows% acc%
2 421 87.89 87.89
[8, 15] 20 4.18 92.07
[16, 31] 30 6.26 98.33
[32, 34] 8 1.67 100.00
Column nonzeros, A|j
range: min A|j: 2 (0.417537%) max A|j: 3 (0.626305%)
distrib: A|j cols cols% acc%
2 435 51.66 51.66
3 407 48.34 100.00
A nonzeros, A(ij)
range: min |A(ij)|: 1.00000 max |A(ij)|: 100.000
distrib: A(ij) coeffs
[1, 10) 1670
[10, 100] 421
-------------------------------------------------------------------------------
Constraint bounds, lb <= Ax <= ub
distrib: |b| lbs ubs
0 421
[1, 10] 58 58
Variable bounds, lb <= x <= ub
0 842
[1, 10) 421
[10, 100] 421
-------------------------------------------------------------------------------
The survey is divided into six dierent sections, each described below. To keep the presentation
short with focus on key elements the analyzer generally attempts to display information on issues
relevant for the current model only: E.g., if the model does not have any conic constraints (this is the
case in the example above) or any integer variables, those parts of the analysis will not appear.
12.1.1 General characteristics
The rst part of the survey consists of a brief summary of the models linear and quadratic constraints
(indexed by i) and variables (indexed by j). The summary is divided into three subsections:
Constraints
upper bd: The number of upper bounded constraints,
n1
j=0
a
ij
x
j
u
c
i
12.1. THE PROBLEM ANALYZER 81
lower bd: The number of lower bounded constraints, l
c
i

n1
j=0
a
ij
x
j
ranged : The number of ranged constraints, l
c
i

n1
j=0
a
ij
x
j
u
c
i
fixed : The number of xed constraints, l
c
i
=
n1
j=0
a
ij
x
j
= u
c
i
free : The number of free constraints
Bounds
upper bd: The number of upper bounded variables, x
j
u
x
j
lower bd: The number of lower bounded variables, l
x
k
x
j
ranged : The number of ranged variables, l
x
k
x
j
u
x
j
fixed : The number of xed variables, l
x
k
= x
j
= u
x
j
free : The number of free variables
Variables
cont: The number of continuous variables, x
j
R
bin : The number of binary variables, x
j
0, 1
int : The number of general integer variables, x
j
Z
Only constraints, bounds and domains actually in the model will be reported on, cf. appendix J; if
all entities in a section turn out to be of the same kind, the number will be replaced by all for brevity.
12.1.2 Objective
The second part of the survey focuses on (the linear part of) the objective, summarizing the opti-
mization sense and the coecients absolute value range and distribution. The number of 0 (zero)
coecients is singled out (if any such variables are in the problem).
The range is displayed using three terms:
min |c|: The minimum absolute value among all coeecients
min |c|>0: The minimum absolute value among the nonzero coecients
max |c|: The maximum absolute value among the coecients
If some of these extrema turn out to be equal, the display is shortened accordingly:
If min |c| is greater than zero, the min |c|>0 term is obsolete and will not be displayed
If only one or two dierent coecients occur this will be displayed using all and an explicit
listing of the coecients
The absolute value distribution is displayed as a table summarizing the numbers by orders of magnitude
(with a ratio of 10). Again, the number of variables with a coecient of 0 (if any) is singled out. Each
line of the table is headed by an interval (half-open intervals including their lower bounds), and is
followed by the number of variables with their objective coecient in this interval. Intervals with no
elements are skipped.
12.1.3 Linear constraints
The third part of the survey displays information on the nonzero coecients of the linear constraint
matrix.
Following a brief summary of the matrix dimensions and the number of nonzero coecients in
total, three sections provide further details on how the nonzero coecients are distributed by row-wise
count (A_i), by column-wise count (A|j), and by absolute value (|A(ij)|). Each section is headed
by a brief display of the distributions range (min and max), and for the row/column-wise counts the
corresponding densities are displayed too (in parentheses).
The distribution tables single out three particularly interesting counts: zero, one, and two nonzeros
per row/column; the remaining row/column nonzeros are displayed by orders of magnitude (ratio 2).
For each interval the relative and accumulated relative counts are also displayed.
Note that constraints may have both linear and quadratic terms, but the empty rows and columns
reported in this part of the survey relate to the linear terms only. If empty rows and/or columns are
found in the linear constraint matrix, the problem is analyzed further in order to determine if the
corresponding constraints have any quadratic terms or the corresponding variables are used in conic
or quadratic constraints; cf. the last two examples of appendix J.
The distribution of the absolute values, |A(ij)|, is displayed just as for the objective coecients
described above.
12.1.4 Constraint and variable bounds
The fourth part of the survey displays distributions for the absolute values of the nite lower and upper
bounds for both constraints and variables. The number of bounds at 0 is singled out and, otherwise,
displayed by orders of magnitude (with a ratio of 10).
12.1.5 Quadratic constraints
The fth part of the survey displays distributions for the nonzero elements in the gradient of the
quadratic constraints, i.e. the nonzero row counts for the column vectors Qx. The table is similar to
the tables for the linear constraints nonzero row and column counts described in the surveys third
part.
Note: Quadratic constraints may also have a linear part, but that will be included in the linear
constraints survey; this means that if a problem has one or more pure quadratic constraints, part three
of the survey will report an equal number of linear constraint rows with 0 (zero) nonzeros, cf. the last
example in appendix J. Likewise, variables that appear in quadratic terms only will be reported as
empty columns (0 nonzeros) in the linear constraint report.
12.2. ANALYZING INFEASIBLE PROBLEMS 83
12.1.6 Conic constraints
The last part of the survey summarizes the models conic constraints. For each of the two types of
cones, quadratic and rotated quadratic, the total number of cones are reported, and the distribution
of the cones dimensions are displayed using intervals. Cone dimensions of 2, 3, and 4 are singled out.
12.2 Analyzing infeasible problems
When developing and implementing a new optimization model, the rst attempts will often be either
infeasible, due to specication of inconsistent constraints, or unbounded, if important constraints have
been left out.
In this chapter we will
go over an example demonstrating how to locate infeasible constraints using the MOSEK infea-
sibility report tool,
discuss in more general terms which properties that may cause infeasibilities, and
present the more formal theory of infeasible and unbounded problems.
12.2.1 Example: Primal infeasibility
A problem is said to be primal infeasible if no solution exists that satisfy all the constraints of the
problem.
As an example of a primal infeasible problem consider the problem of minimizing the cost of trans-
portation between a number of production plants and stores: Each plant produces a xed number of
goods, and each store has a xed demand that must be met. Supply, demand and cost of transportation
per unit are given in gure 12.1.
The problem represented in gure 12.1 is infeasible, since the total demand
2300 = 1100 + 200 + 500 + 500 (12.1)
exceeds the total supply
2200 = 200 + 1000 + 1000 (12.2)
If we denote the number of transported goods from plant i to store j by x
ij
, the problem can be
formulated as the LP:
minimize x
11
+ 2x
12
+ 5x
23
+ 2x
24
+ x
31
+ 2x
33
+ x
34
subject to x
11
+ x
12
200,
x
23
+ x
24
1000,
x
31
+ x
33
+ x
34
1000,
x
11
+ x
31
= 1100,
x
12
= 200,
x
23
+ x
33
= 500,
x
24
+ x
34
= 500,
x
ij
0.
(12.3)
Solving the problem (12.3) using MOSEK will result in a solution, a solution status and a problem
status. Among the log output from the execution of MOSEK on the above problem are the lines:
Supply Demand
1
2
5
2
1
2
1
Plant 1
Plant 2
Plant 3
Store 4
Store 3
Store 2
Store 1
1000
1000
200
500
500
200
1100
Figure 12.1: Supply, demand and cost of transportation.
Basic solution
Problem status : PRIMAL_INFEASIBLE
Solution status : PRIMAL_INFEASIBLE_CER
The rst line indicates that the problem status is primal infeasible. The second line says that a
certicate of the infeasibility was found. The certicate is returned in place of the solution to the
problem.
12.2.2 Locating the cause of primal infeasibility
Usually a primal infeasible problem status is caused by a mistake in formulating the problem and
therefore the question arises: What is the cause of the infeasible status? When trying to answer this
question, it is often advantageous to follow these steps:
Remove the objective function. This does not change the infeasible status but simplies the
problem, eliminating any possibility of problems related to the objective function.
Consider whether your problem has some necessary conditions for feasibility and examine if these
are satised, e.g. total supply should be greater than or equal to total demand.
Verify that coecients and bounds are reasonably sized in your problem.
If the problem is still primal infeasible, some of the constraints must be relaxed or removed completely.
The MOSEK infeasibility report (Section 12.2.4) may assist you in nding the constraints causing the
infeasibility.
Possible ways of relaxing your problem include:
Increasing (decreasing) upper (lower) bounds on variables and constraints.
Removing suspected constraints from the problem.
Returning to the transportation example, we discover that removing the fth constraint
x
12
= 200 (12.4)
makes the problem feasible.
12.2.3 Locating the cause of dual infeasibility
A problem may also be dual infeasible. In this case the primal problem is often unbounded, mening
that feasbile solutions exists such that the objective tends towards innity. An example of a dual
infeasible and primal unbounded problem is:
minimize x
1
subject to x
1
5.
(12.5)
To resolve a dual infeasibility the primal problem must be made more restricted by
Adding upper or lower bounds on variables or constraints.
Removing variables.
Changing the objective.
12.2.3.1 A cautious note
The problem
minimize 0
subject to 0 x
1
,
x
j
x
j+1
, j = 1, . . . , n 1,
x
n
1
(12.6)
is clearly infeasible. Moreover, if any one of the constraints are dropped, then the problem becomes
feasible.
This illustrates the worst case scenario that all, or at least a signicant portion, of the constraints
are involved in the infeasibility. Hence, it may not always be easy or possible to pinpoint a few
constraints which are causing the infeasibility.
12.2.4 The infeasibility report
MOSEK includes functionality for diagnosing the cause of a primal or a dual infeasibility. It can be
turned on by setting the MSK IPAR INFEAS REPORT AUTO to MSK ON. This causes MOSEK to print a
report on variables and constraints involved in the infeasibility.
The MSK IPAR INFEAS REPORT LEVEL parameter controls the amount of information presented in
the infeasibility report. The default value is 1.
12.2.4.1 Example: Primal infeasibility
We will reuse the example (12.3) located in infeas.lp:
\
\ An example of an infeasible linear problem.
\
minimize
obj: + 1 x11 + 2 x12 + 1 x13
+ 4 x21 + 2 x22 + 5 x23
+ 4 x31 + 1 x32 + 2 x33
st
s0: + x11 + x12 <= 200
s1: + x23 + x24 <= 1000
s2: + x31 +x33 + x34 <= 1000
d1: + x11 + x31 = 1100
d2: + x12 = 200
d3: + x23 + x33 = 500
d4: + x24 + x34 = 500
bounds
end
Using the command line
mosek -d MSK_IPAR_INFEAS_REPORT_AUTO MSK_ON infeas.lp
MOSEK produces the following infeasibility report
MOSEK PRIMAL INFEASIBILITY REPORT.
Problem status: The problem is primal infeasible
The following constraints are involved in the primal infeasibility.
Index Name Lower bound Upper bound Dual lower Dual upper
0 s0 NONE 2.000000e+002 0.000000e+000 1.000000e+000
2 s2 NONE 1.000000e+003 0.000000e+000 1.000000e+000
3 d1 1.100000e+003 1.100000e+003 1.000000e+000 0.000000e+000
4 d2 2.000000e+002 2.000000e+002 1.000000e+000 0.000000e+000
The following bound constraints are involved in the infeasibility.
Index Name Lower bound Upper bound Dual lower Dual upper
8 x33 0.000000e+000 NONE 1.000000e+000 0.000000e+000
10 x34 0.000000e+000 NONE 1.000000e+000 0.000000e+000
The infeasibility report is divided into two sections where the rst section shows which constraints that
are important for the infeasibility. In this case the important constraints are the ones named s0, s2, d1,
and d2. The values in the columns Dual lower and Dual upper are also useful, since a non-zero
dual lower value for a constraint implies that the lower bound on the constraint is important for the
infeasibility. Similarly, a non-zero dual upper value implies that the upper bound on the constraint is
important for the infeasibility.
It is also possible to obtain the infeasible subproblem. The command line
mosek -d MSK_IPAR_INFEAS_REPORT_AUTO MSK_ON infeas.lp -info rinfeas.lp
produces the les rinfeas.bas.inf.lp. In this case the content of the le rinfeas.bas.inf.lp is
minimize
Obj: + CFIXVAR
st
s0: + x11 + x12 <= 200
s2: + x31 + x33 + x34 <= 1e+003
d1: + x11 + x31 = 1.1e+003
d2: + x12 = 200
bounds
x11 free
x12 free
x13 free
x21 free
x22 free
x23 free
x31 free
x32 free
x24 free
CFIXVAR = 0e+000
end
which is an optimization problem. This problem is identical to (12.3), except that the objective and
some of the constraints and bounds have been removed. Executing the command
mosek -d MSK_IPAR_INFEAS_REPORT_AUTO MSK_ON rinfeas.bas.inf.lp
demonstrates that the reduced problem is primal infeasible. Since the reduced problem is usually
smaller than original problem, it should be easier to locate the cause of the infeasibility in this rather
than in the original (12.3).
12.2.4.2 Example: Dual infeasibility
The example problem
maximize - 200 y1 - 1000 y2 - 1000 y3
- 1100 y4 - 200 y5 - 500 y6
- 500 y7
subject to
x11: y1+y4 < 1
x12: y1+y5 < 2
x23: y2+y6 < 5
x24: y2+y7 < 2
x31: y3+y4 < 1
x33: y3+y6 < 2
x44: y3+y7 < 1
bounds
y1 < 0
y2 < 0
y3 < 0
y4 free
y5 free
y6 free
y7 free
end
is dual infeasible. This can be veried by proving that
y1=-1, y2=-1, y3=0, y4=1, y5=1
is a certicate of dual infeasibility. In this example the following infeasibility report is produced
(slightly edited):
The following constraints are involved in the infeasibility.
Index Name Activity Objective Lower bound Upper bound
0 x11 -1.000000e+00 NONE 1.000000e+00
4 x31 -1.000000e+00 NONE 1.000000e+00
The following variables are involved in the infeasibility.
Index Name Activity Objective Lower bound Upper bound
3 y4 -1.000000e+00 -1.100000e+03 NONE NONE
Interior-point solution
Problem status : DUAL_INFEASIBLE
Solution status : DUAL_INFEASIBLE_CER
Primal - objective: 1.1000000000e+03 eq. infeas.: 0.00e+00 max bound infeas.: 0.00e+00 cone infeas.: 0.00e+00
Dual - objective: 0.0000000000e+00 eq. infeas.: 0.00e+00 max bound infeas.: 0.00e+00 cone infeas.: 0.00e+00
Let x
denote the reported primal solution. MOSEK states

that the problem is dual infeasible,
that the reported solution is a certicate of dual infeasibility, and
that the infeasibility measure for x
is approximately zero.
Since it was an maximization problem, this implies that
c
t
x
> 0. (12.7)
For a minimization problem this inequality would have been reversed see (12.19).
From the infeasibility report we see that the variable y4, and the constraints x11 and x33 are
involved in the infeasibility since these appear with non-zero values in the Activity column.
One possible strategy to x the infeasibility is to modify the problem so that the certicate of
infeasibility becomes invalid. In this case we may do one the the following things:
Put a lower bound in y3. This will directly invalidate the certicate of dual infeasibility.
Increase the object coecient of y3. Changing the coecients suciently will invalidate the
inequality (12.7) and thus the certicate.
Put lower bounds on x11 or x31. This will directly invalidate the certicate of infeasibility.
Please note that modifying the problem to invalidate the reported certicate does not imply that the
problem becomes dual feasible the infeasibility may simply move, resulting in a new infeasibility.
More often, the reported certicate can be used to give a hint about errors or inconsistencies in
the model that produced the problem.
12.2.5 Theory concerning infeasible problems
This section discusses the theory of infeasibility certicates and how MOSEK uses a certicate to
produce an infeasibility report. In general, MOSEK solves the problem
minimize c
T
x + c
f
subject to l
c
Ax u
c
,
l
x
x u
x
(12.8)
where the corresponding dual problem is
maximize (l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+(l
x
)
T
s
x
l
(u
x
)
T
s
x
u
+ c
f
subject to A
T
y + s
x
l
s
x
u
= c,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
(12.9)
We use the convension that for any bound that is not nite, the corresponding dual variable is xed
at zero (and thus will have no inuence on the dual problem). For example
l
x
j
= (s
x
l
)
j
= 0 (12.10)
12.2.6 The certicate of primal infeasibility
A certicate of primal infeasibility is any solution to the homogenized dual problem
maximize (l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+(l
x
)
T
s
x
l
(u
x
)
T
s
x
u
subject to A
T
y + s
x
l
s
x
u
= 0,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
(12.11)
with a positive objective value. That is, (s
c
l
, s
c
u
, s
x
l
, s
x
u
) is a certicate of primal infeasibility if
(l
c
)
T
s
c
l
(u
c
)
T
s
c
u
+ (l
x
)
T
s
x
l
(u
x
)
T
s
x
u
> 0 (12.12)
and
A
T
y + s
x
l
s
x
u
= 0,
y + s
c
l
s
c
u
= 0,
s
c
l
, s
c
u
, s
x
l
, s
x
u
0.
(12.13)
The well-known Farkas Lemma tells us that (12.8) is infeasible if and only if a certicate of primal
infeasibility exists.
Let (s
c
l
, s
c
u
, s
x
l
, s
x
u
) be a certicate of primal infeasibility then
(s
c
l
)
i
> 0 ((s
c
u
)
i
> 0) (12.14)
implies that the lower (upper) bound on the ith constraint is important for the infeasibility. Further-
more,
(s
x
l
)
j
> 0 ((s
x
u
)
i
> 0) (12.15)
implies that the lower (upper) bound on the jth variable is important for the infeasibility.
12.2.7 The certicate of dual infeasibility
A certicate of dual infeasibility is any solution to the problem
minimize c
T
x
subject to

l
c
Ax u
c
,
l
x
x u
x
(12.16)
with negative objective value, where we use the denitions
l
c
i
:=
_
0, l
c
i
> ,
, otherwise,
u
c
i
:=
_
0, u
c
i
< ,
, otherwise,
(12.17)
and
l
x
i
:=
_
0, l
x
i
> ,
, otherwise,
and u
x
i
:=
_
0, u
x
i
< ,
, otherwise.
(12.18)
Stated dierently, a certicate of dual infeasibility is any x
such that
c
T
x
< 0,
l
c
Ax
u
c
,
l
x
x
u
x
(12.19)
The well-known Farkas Lemma tells us that (12.9) is infeasible if and only if a certicate of dual
infeasibility exists.
Note that if x
is a certicate of dual infeasibility then for any j such that

x
j
,= 0, (12.20)
variable j is involved in the dual infeasibility.
Chapter 13
Feasibility repair
Section 12.2.2 discusses how MOSEK treats infeasible problems. In particular, it is discussed which
information MOSEK returns when a problem is infeasible and how this information can be used to
pinpoint the elements causing the infeasibility.
In this section we will discuss a method for repairing a primal infeasible problem by relaxing the
constraints in a controlled way. For the sake of simplicity we discuss the method in the context
of linear optimization. MOSEK can also repair infeasibilities in quadratic and conic optimization
problems possibly having integer constrained variables. Please note that infeasibilities in nonlinear
optimization problems cant be repaired using the method described below.
13.1 The main idea
Consider the linear optimization problem with m constraints and n variables
minimize c
T
x + c
f
subject to l
c
Ax u
c
,
l
x
x u
x
,
(13.1)
which we assume is infeasible. Moreover, we assume that
(l
c
)
i
(u
c
)
i
, i (13.2)
and
(l
x
)
j
(u
x
)
j
, j (13.3)
because otherwise the problem (13.1) is trivially infeasible.
One way of making the problem feasible is to reduce the lower bounds and increase the upper
bounds. If the change is suciently large the problem becomes feasible.
One obvious question is: What is the smallest change to the bounds that will make the problem
feasible?
We associate a weight with each bound:
w
c
l
R
m
(associated with l
c
),
91
92 CHAPTER 13. FEASIBILITY REPAIR
w
c
u
R
m
(associated with u
c
),
w
x
l
R
n
(associated with l
x
),
w
x
u
R
n
(associated with u
x
),
Now, the problem
minimize p
subject to l
c
Ax + v
c
l
v
c
u
u
c
,
l
x
x + v
x
l
v
x
u
u
x
,
(w
c
l
)
T
v
c
l
+ (w
c
u
)
T
v
c
u
+ (w
x
l
)
T
v
x
l
+ (w
x
u
)
T
v
x
u
p 0,
v
c
l
, v
c
u
, v
x
l
, v
x
u
0
(13.4)
minimizes the weighted sum of changes to the bounds that makes the problem feasible. The variables
(v
c
l
)
i
, (v
c
u
)
i
, (v
x
l
)
i
and (v
c
u
)
i
are elasticity variables because they allow a constraint to be violated and
hence add some elasticity to the problem. For instance, the elasticity variable (v
c
l
)
i
shows how much
the lower bound (l
c
)
i
should be relaxed to make the problem feasible. Since p is minimized and
(w
c
l
)
T
v
c
l
+ (w
c
u
)
T
v
c
u
+ (w
x
l
)
T
v
x
l
+ (w
x
u
)
T
v
x
u
p 0, (13.5)
a large (w
c
l
)
i
tends to imply that the elasticity variable (v
c
l
)
i
will be small in an optimal solution.
The reader may want to verify that the problem (13.4) is always feasible given the assumptions
(13.2) and (13.3).
Please note that if a weight is negative then the resulting problem (13.4) is unbounded.
The weights w
c
l
, w
c
u
, w
x
l
, and w
x
u
can be regarded as a costs (penalties) for violating the associated
constraints. Thus a higher weight implies that higher priority is given to the satisfaction of the
associated constraint.
The main idea can now be presented as follows. If you have an infeasible problem, then form
the problem (13.4) and optimize it. Next inspect the optimal solution (v
c
l
)
, (v
c
u
)
, (v
x
l
)
, and (v
x
u
)
to problem (13.4). This solution provides a suggested relaxation of the bounds that will make the
problem feasible.
Assume that p
is an optimal objective value to (13.4). An extension of the idea presented above

is to solve the problem
minimize c
T
x
subject to l
c
Ax + v
c
l
v
c
u
u
c
,
l
x
x + v
x
l
v
x
u
u
x
,
(w
c
l
)
T
v
c
l
+ (w
c
u
)
T
v
c
u
+ (w
x
l
)
T
v
x
l
+ (w
x
u
)
T
v
x
u
p 0,
p = p
,
v
c
l
, v
c
u
, v
x
l
, v
x
u
0
(13.6)
which minimizes the true objective while making sure that total weighted violations of the bounds is
minimal, i.e. equals to p
.
13.2 Feasibility repair in MOSEK
MOSEK includes functionality that help you construct the problem (13.4) simply by passing a set of
weights to MOSEK. This can be used for linear, quadratic, and conic optimization problems, possibly
having integer constrained variables.
13.2. FEASIBILITY REPAIR IN MOSEK 93
13.2.1 Usage of negative weights
As the problem (13.4) is presented it does not make sense to use negative weights since that makes
the problem unbounded. Therefore, if the value of a weight is negative MOSEK xes the associated
elasticity variable to zero, e.g. if
(w
c
l
)
i
< 0
then MOSEK imposes the bound
(v
c
l
)
i
0.
This implies that the lower bound on the ith constraint will not be violated. (Clearly, this could
also imply that the problem is infeasible so negative weight should be used with care). Associating a
negative weights with a constraint tells MOSEK that the constraint should not be relaxed.
13.2.2 Automatical naming
MOSEK can automatically create a new problem of the form (13.4) starting from an existing problem
by adding the elasticity variables and the extra constraints. Specicly, the variables v
c
l
, v
c
u
, v
x
l
, v
x
u
, and
p are appended to existing variable vector x in their natural order. Moreover, the constraint (13.5) is
appended to the constraints.
The new variables and constraints are automatically given names as follows:
The names of the variables (v
c
l
)
i
and (v
c
u
)
i
are constructed from the name of the ith constraint.
For instance, if the 9th original constraint is named c9, then by default (v
c
l
)
9
and (v
c
u
)
9
are given
the names LO*c9 and UP*c9 respectively. If necessary, the character * can be replaced by a
dierent string by changing the
MSK SPAR FEASREPAIR NAME SEPARATOR
parameter.
The additional constraints
l
x
x + v
x
l
v
x
u
u
x
are given names as follows. There is exactly one constraint per variable in the original problem,
and thus the ith of these constraints is named after the ith variable in the original problem.
For instance, if the rst original variable is named x0, then the rst of the above constraints
is named MSK-x1. If necessary, the prex MSK- can be replaced by a dierent string by
changing the
MSK SPAR FEASREPAIR NAME PREFIX
parameter.
The variable p is by default given the name WSUMVIOLVAR, and the constraint (13.5) is given the
name WSUMVIOLCON.
The substring WSUMVIOL can be replaced by a dierent string by changing the
MSK SPAR FEASREPAIR NAME WSUMVIOL
parameter.
94 CHAPTER 13. FEASIBILITY REPAIR
13.2.3 An example
Consider the example linear optimization
minimize 10x
1
9x
2
,
subject to 7/10x
1
+ 1x
2
630,
1/2x
1
+ 5/6x
2
600,
1x
1
+ 2/3x
2
708,
1/10x
1
+ 1/4x
2
135,
x
1
, x
2
0.
x
2
650
(13.7)
This is an infeasible problem. Now suppose we wish to use MOSEK to suggest a modication to the
bounds that makes the problem feasible.
The command
mosek -d MSK_IPAR_FEASREPAIR_OPTIMIZE
MSK_FEASREPAIR_OPTIMIZE_PENALTY -d
MSK_IPAR_OPF_WRITE_SOLUTIONS MSK_ON feasrepair.lp
-infrepo minv.opf
writes the problem (13.4) and its solution to an OPF formatted le. In this case the le minv.opf.
The parameter
MSK_IPAR_FEASREPAIR_OPTIMIZE
controls whether the function returns the problem (13.4) or the problem (13.6). In the case
MSK_IPAR_FEASREPAIR_OPTIMIZE
is equal to
MSK_FEASREPAIR_OPTIMIZE_NONE
then (13.4) is returned, but the problem is not solved. For MSK FEASREPAIR OPTIMIZE PENALTY the
problem (13.4) is returned and solved. Finally for MSK FEASREPAIR OPTIMIZE COMBINED (13.6) is
returned and solved.
Chapter 14
Sensitivity analysis
14.1 Introduction
Given an optimization problem it is often useful to obtain information about how the optimal objective
value change when the problem parameters are perturbed. For instance assume that a bound represents
a capacity of a machine. Now, it may be possible to expand the capacity for a certain cost and hence
it worthwhile to know what the value of additional capacity is. This is precisely the type of questions
sensitivity analysis deals with.
Analyzing how the optimal objective value changes when the problem data is changed is called
sensitivity analysis.
14.2 Restrictions
Currently, sensitivity analysis is only available for continuous linear optimization problems. Moreover,
MOSEK can only deal with perturbations in bounds or objective coecients.
14.3 References
The book [12] discusses the classical sensitivity analysis in Chapter 10 whereas the book [19, Chapter
19] presents a modern introduction to sensitivity analysis. Finally, it is recommended to read the short
paper [21] to avoid some of the pitfalls associated with sensitivity analysis.
14.4 Sensitivity analysis for linear problems
14.4.1 The optimal objective value function
Assume that we are given the problem
z(l
c
, u
c
, l
x
, u
x
, c) = minimize c
T
x
subject to l
c
Ax u
c
,
l
x
x u
x
,
(14.1)
95
96 CHAPTER 14. SENSITIVITY ANALYSIS
and we want to know how the optimal objective value changes as l
c
i
is perturbed. In order to answer
this question then dene the perturbed problem for l
c
i
as follows
f
l
c
i
() = minimize c
T
x
subject to l
c
+ e
i
Ax u
c
,
l
x
x u
x
,
(14.2)
where e
i
is the ith column of the identity matrix. The function
f
l
c
i
() (14.3)
shows the optimal objective value as a function of . Note that a change in corresponds to a
perturbation in l
c
i
and hence (14.3) shows the optimal objective value as a function of l
c
i
.
It is possible to prove that the function (14.3) is a piecewise linear and convex function i.e. the
function may look like the illustration in Figure 14.1.
f( )
0
1
2
f( )
0

2 1
Figure 14.1: The optimal value function f
l
c
i
(). Left: = 0 is in the interior of linearity interval.
Right: = 0 is a breakpoint.
Clearly, if the function f
l
c
i
() does not change much when is changed, then we can conclude that
the optimal objective value is insensitive to changes in l
c
i
. Therefore, we are interested in how f
l
c
i
()
changes for small changes in . Now dene
f
l
c
i
(0) (14.4)
to be the so called shadow price related to l
c
i
. The shadow price species how the objective value
changes for small changes in around zero. Moreover, we are interested in the so called linearity
interval
[
1
,
2
] (14.5)
for which
f
l
c
i
() = f
l
c
i
(0). (14.6)
To summarize the sensitivity analysis provides a shadow price and the linearity interval in which
the shadow price is constant.
The reader may have noticed that we are sloppy in the denition of the shadow price. The reason
is that the shadow price is not dened in the right example in Figure 14.1 because the function f
l
c
i
()
is not dierentiable for = 0. However, in that case we can dene a left and a right shadow price and
a left and a right linearity interval.
14.4. SENSITIVITY ANALYSIS FOR LINEAR PROBLEMS 97
In the above discussion we only discussed changes in l
c
i
. We dene the other optimal objective
value functions as follows
f
u
c
i
() = z(l
c
, u
c
+ e
i
, l
x
, u
x
, c), i = 1, . . . , m,
f
l
x
j
() = z(l
c
, u
c
, l
x
+ e
j
, u
x
, c), j = 1, . . . , n,
f
u
x
j
() = z(l
c
, u
c
, l
x
, u
x
+ e
j
, c), j = 1, . . . , n,
f
cj
() = z(l
c
, u
c
, l
x
, u
x
, c + e
j
), j = 1, . . . , n.
(14.7)
Given these denitions it should be clear how linearity intervals and shadow prices are dened for the
parameters u
c
i
etc.
14.4.1.1 Equality constraints
In MOSEK a constraint can be specied as either an equality constraints or a ranged constraints.
Suppose constraint i is an equality constraint. We then dene the optimal value function for constraint
i by
f
e
c
i
() = z(l
c
+ e
i
, u
c
+ e
i
, l
x
, u
x
, c) (14.8)
Thus for a equality constraint the upper and lower bound (which are equal) are perturbed simultane-
ously. From the point of view of MOSEK sensitivity analysis a ranged constrain with l
c
i
= u
c
i
therefore
diers from an equality constraint.
14.4.2 The basis type sensitivity analysis
The classical sensitivity analysis discussed in most textbooks about linear optimization, e.g. [12,
Chapter 10], is based on an optimal basic solution or equivalently on an optimal basis. This method
may produce misleading results [19, Chapter 19] but is computationally cheap. Therefore, and for
historical reasons this method is available in MOSEK.
We will now briey discuss the basis type sensitivity analysis. Given an optimal basic solution
which provides a partition of variables into basic and non-basic variables then the basis type sensitivity
analysis computes the linearity interval [
1
,
2
] such that the basis remains optimal for the perturbed
problem. A shadow price associated with the linearity interval is also computed. However, it is well-
known that an optimal basic solution may not be unique and therefore the result depends on the
optimal basic solution employed in the sensitivity analysis. This implies that the computed interval is
only a subset of the largest interval for which the shadow price is constant. Furthermore, the optimal
objective value function might have a breakpoint for = 0. In this case the basis type sensitivity
method will only provide a subset of either the left or the right linearity interval.
In summary the basis type sensitivity analysis is computationally cheap but does not provide
complete information. Hence, the results of the basis type sensitivity analysis should be used with
care.
14.4.3 The optimal partition type sensitivity analysis
Another method for computing the complete linearity interval is called the optimal partition type
sensitivity analysis. The main drawback to the optimal partition type sensitivity analysis is it is
computationally expensive. This type of sensitivity analysis is currently provided as an experimental
feature in MOSEK.
Given optimal primal and dual solutions to (14.1) i.e. x
and ((s
c
l
)
, (s
c
u
)
, (s
x
l
)
, (s
x
u
)
) then the
optimal objective value is given by
z
:= c
T
x
. (14.9)
The left and right shadow prices
1
and
2
for l
c
i
is given by the pair of optimization problems
1
= minimize e
T
i
s
c
l
subject to A
T
(s
c
l
s
c
u
) + s
x
l
s
x
u
= c,
(l
c
)
T
(s
c
l
) (u
c
)
T
(s
c
u
) + (l
x
)
T
(s
x
l
) (u
x
)
T
(s
x
u
) = z
,
s
c
l
, s
c
u
, s
c
l
, s
x
u
0
(14.10)
and
2
= maximize e
T
i
s
c
l
subject to A
T
(s
c
l
s
c
u
) + s
x
l
s
x
u
= c,
(l
c
)
T
(s
c
l
) (u
c
)
T
(s
c
u
) + (l
x
)
T
(s
x
l
) (u
x
)
T
(s
x
u
) = z
,
s
c
l
, s
c
u
, s
c
l
, s
x
u
0.
(14.11)
The above two optimization problems makes it easy to interpret-ate the shadow price. Indeed assume
that ((s
c
l
)
, (s
c
u
)
, (s
x
l
)
, (s
x
u
)
) is an arbitrary optimal solution then it must hold

(s
c
l
)
i
[
1
,
2
]. (14.12)
Next the linearity interval [
1
,
2
] for l
c
i
is computed by solving the two optimization problems
1
= minimize
subject to l
c
+ e
i
Ax u
c
,
c
T
x
1
= z
,
l
x
x u
x
,
(14.13)
and
2
= maximize
subject to l
c
+ e
i
Ax u
c
,
c
T
x
2
= z
,
l
x
x u
x
.
(14.14)
The linearity intervals and shadow prices for u
c
i
, l
x
j
, and u
x
j
can be computed in a similar way to
how it is computed for l
c
i
.
The left and right shadow price for c
j
denoted
1
and
2
respectively is given by the pair optimiza-
tion problems
1
= minimize e
T
j
x
subject to l
c
+ e
i
Ax u
c
,
c
T
x = z
,
l
x
x u
x
(14.15)
and
2
= maximize e
T
j
x
subject to l
c
+ e
i
Ax u
c
,
c
T
x = z
,
l
x
x u
x
.
(14.16)
Once again the above two optimization problems makes it easy to interpret-ate the shadow prices.
Indeed assume that x
is an arbitrary primal optimal solution then it must hold

x
j
[
1
,
2
]. (14.17)
The linearity interval [
1
,
2
] for a c
j
is computed as follows
1
= minimize
subject to A
T
(s
c
l
s
c
u
) + s
x
l
s
x
u
= c + e
j
,
(l
c
)
T
(s
c
l
) (u
c
)
T
(s
c
u
) + (l
x
)
T
(s
x
l
) (u
x
)
T
(s
x
u
)
1
z
,
s
c
l
, s
c
u
, s
c
l
, s
x
u
0
(14.18)
and
2
= maximize
subject to A
T
(s
c
l
s
c
u
) + s
x
l
s
x
u
= c + e
j
,
(l
c
)
T
(s
c
l
) (u
c
)
T
(s
c
u
) + (l
x
)
T
(s
x
l
) (u
x
)
T
(s
x
u
)
2
z
,
s
c
l
, s
c
u
, s
c
l
, s
x
u
0.
(14.19)
14.4.4 An example
As an example we will use the following transportation problem. Consider the problem of minimizing
the transportation cost between a number of production plants and stores. Each plant supplies a
number of goods and each store has a given demand that must be met. Supply, demand and cost of
transportation per unit are shown in Figure 14.2.
If we denote the number of transported goods from location i to location j by x
ij
, the problem can
be formulated as the linear optimization problem
minimize
1x
11
+ 2x
12
+ 5x
23
+ 2x
24
+ 1x
31
+ 2x
33
+ 1x
34
(14.20)
subject to
x
11
+ x
12
400,
x
23
+ x
24
1200,
x
31
+ x
33
+ x
34
1000,
x
11
+ x
31
= 800,
x
12
= 100,
x
23
+ x
33
= 500,
x
24
+ x
34
= 500,
x
11
, x
12
, x
23
, x
24
, x
31
, x
33
, x
34
0.
(14.21)
The basis type and the optimal partition type sensitivity results for the transportation problem is
shown in Table 14.1 and 14.2 respectively.
Looking at the results from the optimal partition type sensitivity analysis we see that for the
constraint number 1 we have
1
,=
2
and
1
,=
2
. Therefore, we have a left linearity interval of
[300, 0] and a right interval of [0, 500]. The corresponding left and right shadow price is 3 and 1
respectively. This implies that if the upper bound on constraint 1 increases by
[0,
1
] = [0, 500] (14.22)
Supply Demand
1
2
5
2
1
2
1
Plant 1
Plant 2
Plant 3
Store 4
Store 3
Store 2
Store 1
1000
500
500
800
100
400
1200
Figure 14.2: Supply, demand and cost of transportation.
Basis type
Con.
1

2

1

2
1 300.00 0.00 3.00 3.00
2 700.00 + 0.00 0.00
3 500.00 0.00 3.00 3.00
4 0.00 500.00 4.00 4.00
5 0.00 300.00 5.00 5.00
6 0.00 700.00 5.00 5.00
7 500.00 700.00 2.00 2.00
Var.
1

2

1

2
x
11
300.00 0.00 0.00
x
12
100.00 0.00 0.00
x
23
0.00 0.00 0.00
x
24
500.00 0.00 0.00
x
31
500.00 0.00 0.00
x
33
500.00 0.00 0.00
x
34
0.000000 500.00 2.00 2.00
Optimal partition type
Con.
1

2

1

2
1 300.00 500.00 3.00 1.00
2 700.00 + 0.00 0.00
3 500.00 500.00 3.00 1.00
4 500.00 500.00 2.00 4.00
5 100.00 300.00 3.00 5.00
6 500.00 700.00 3.00 5.00
7 500.00 700.00 2.00 2.00
Var.
1

2

1

2
x
11
300.00 0.00 0.00
x
12
100.00 0.00 0.00
x
23
500.00 0.00 2.00
x
24
500.00 0.00 0.00
x
31
500.00 0.00 0.00
x
33
500.00 0.00 0.00
x
34
500.00 0.00 2.00
Table 14.1: Ranges and shadow prices related to bounds on constraints and variables. Left: Results
for basis type sensitivity analysis. Right: Results for the optimal partition type sensitivity analysis.
Basis type
Var.
1

2

1

2
c
1
3.00 300.00 300.00
c
2
100.00 100.00
c
3
2.00 0.00 0.00
c
4
2.00 500.00 500.00
c
5
3.00 500.00 500.00
c
6
2.00 500.00 500.00
c
7
2.00 0.00 0.00
Optimal partition type
Var.
1

2

1

2
c
1
3.00 300.00 300.00
c
2
100.00 100.00
c
3
2.00 0.00 0.00
c
4
2.00 500.00 500.00
c
5
3.00 500.00 500.00
c
6
2.00 500.00 500.00
c
7
2.00 0.00 0.00
Table 14.2: Ranges and shadow prices related to the objective coecients. Left: Results for basis type
sensitivity analysis. Right: Results for the optimal partition type sensitivity analysis.
then the optimal objective value will decrease by the value
2
= 1. (14.23)
Correspondingly, if the upper bound on constraint 1 is decreased by
[0, 300] (14.24)
then the optimal objective value will increased by the value
1
= 3. (14.25)
14.5 Sensitivity analysis with the command line tool
A sensitivity analysis can be performed with the MOSEK command line tool using the command
mosek myproblem.mps -sen sensitivity.ssp
where sensitivity.ssp is a le in the format described in the next section. The ssp le describes
which parts of the problem the sensitivity analysis should be performed on.
By default results are written to a le named myproblem.sen. If necessary, this lename can be
changed by setting the
MSK SPAR SENSITIVITY RES FILE NAME
parameter By default a basis type sensitivity analysis is performed. However, the type of sensitivity
analysis (basis or optimal partition) can be changed by setting the parameter
MSK IPAR SENSITIVITY TYPE
appropriately. Following values are accepted for this parameter:
MSK SENSITIVITY TYPE BASIS
MSK SENSITIVITY TYPE OPTIMAL PARTITION
It is also possible to use the command line
mosek myproblem.mps -d MSK_IPAR_SENSITIVITY_ALL MSK_ON
in which case a sensitivity analysis on all the parameters is performed.
14.5.1 Sensitivity analysis specication le
MOSEK employs an MPS like le format to specify on which model parameters the sensitivity anal-
ysis should be performed. As the optimal partition type sensitivity analysis can be computationally
expensive it is important to limit the sensitivity analysis.
The format of the sensitivity specication le is shown in gure 14.3, where capitalized names are
keywords, and names in brackets are names of the constraints and variables to be included in the
analysis.
The sensitivity specication le has three sections, i.e.
BOUNDS CONSTRAINTS: Species on which bounds on constraints the sensitivity analysis should
be performed.
14.5. SENSITIVITY ANALYSIS WITH THE COMMAND LINE TOOL 103
* A comment
BOUNDS CONSTRAINTS
U|L|LU [cname1]
U|L|LU [cname2]-[cname3]
BOUNDS VARIABLES
U|L|LU [vname1]
U|L|LU [vname2]-[vname3]
OBJECTIVE VARIABLES
[vname1]
[vname2]-[vname3]
Figure 14.3: The sensitivity analysis le format.
BOUNDS VARIABLES: Species on which bounds on variables the sensitivity analysis should be
performed.
OBJECTIVE VARIABLES: Species on which objective coecients the sensitivity analysis should
be performed.
A line in the body of a section must begin with a whitespace. In the BOUNDS sections one of the keys
L, U, and LU must appear next. These keys specify whether the sensitivity analysis is performed on
the lower bound, on the upper bound, or on both the lower and the upper bound respectively. Next,
a single constraint (variable) or range of constraints (variables) is specied.
Recall from Section 14.4.1.1 that equality constraints are handled in a special way. Sensitivity
analysis of an equality constraint can be specied with either L, U, or LU, all indicating the same,
namely that upper and lower bounds (which are equal) are perturbed simultaneously.
As an example consider
BOUNDS CONSTRAINTS
L "cons1"
U "cons2"
LU "cons3"-"cons6"
which requests that sensitivity analysis is performed on the lower bound of the constraint named
cons1, on the upper bound of the constraint named cons2, and on both lower and upper bound on
the constraints named cons3 to cons6.
It is allowed to use indexes instead of names, for instance
BOUNDS CONSTRAINTS
L "cons1"
U 2
LU 3 - 6
The character * indicates that the line contains a comment and is ignored.
14.5.2 Example: Sensitivity analysis from command line
As an example consider the sensitivity.ssp le shown in Figure 14.4.
The command
* Comment 1
BOUNDS CONSTRAINTS
U "c1" * Analyze upper bound for constraint named c1
U 2 * Analyze upper bound for the second constraint
U 3-5 * Analyze upper bound for constraint number 3 to number 5
BOUNDS VARIABLES
L 2-4 * This section specifies which bounds on variables should be analyzed
L "x11"
OBJECTIVE VARIABLES
"x11" * This section specifies which objective coefficients should be analyzed
2
Figure 14.4: Example of the sensitivity le format.
mosek transport.lp -sen sensitivity.ssp -d MSK_IPAR_SENSITIVITY_TYPE MSK_SENSITIVITY_TYPE_BASIS
produces the transport.sen le shown below.
BOUNDS CONSTRAINTS
INDEX NAME BOUND LEFTRANGE RIGHTRANGE LEFTPRICE RIGHTPRICE
0 c1 UP -6.574875e-18 5.000000e+02 1.000000e+00 1.000000e+00
2 c3 UP -6.574875e-18 5.000000e+02 1.000000e+00 1.000000e+00
3 c4 FIX -5.000000e+02 6.574875e-18 2.000000e+00 2.000000e+00
4 c5 FIX -1.000000e+02 6.574875e-18 3.000000e+00 3.000000e+00
5 c6 FIX -5.000000e+02 6.574875e-18 3.000000e+00 3.000000e+00
BOUNDS VARIABLES
INDEX NAME BOUND LEFTRANGE RIGHTRANGE LEFTPRICE RIGHTPRICE
2 x23 LO -6.574875e-18 5.000000e+02 2.000000e+00 2.000000e+00
3 x24 LO -inf 5.000000e+02 0.000000e+00 0.000000e+00
4 x31 LO -inf 5.000000e+02 0.000000e+00 0.000000e+00
0 x11 LO -inf 3.000000e+02 0.000000e+00 0.000000e+00
OBJECTIVE VARIABLES
INDEX NAME LEFTRANGE RIGHTRANGE LEFTPRICE RIGHTPRICE
0 x11 -inf 1.000000e+00 3.000000e+02 3.000000e+02
2 x23 -2.000000e+00 +inf 0.000000e+00 0.000000e+00
14.5.3 Controlling log output
Setting the parameter
MSK_IPAR_LOG_SENSITIVITY
to 1 or 0 (default) controls whether or not the results from sensitivity calculations are printed to the
message stream.
The parameter
MSK_IPAR_LOG_SENSITIVITY_OPT
controls the amount of debug information on internal calculations from the sensitivity analysis.
Appendix A
MOSEK command line tool
reference
A.1 Introduction
The MOSEK command line tool is used to solve optimization problems from the operating system
command line. It is invoked as follows
mosek [options] [filename]
where both [options] and [filename] are optional arguments. [filename] is a le describing the
optimization problems and is either a MPS le or AMPL nl le. [options] consists of command line
arguments that modies the behavior of MOSEK.
A.2 Command line arguments
The following list shows the possible command-line arguments for MOSEK:
-a MOSEK runs in AMPL mode.
-AMPL The input le is an AMPL nl le.
-basi name Input basis solution le name.
-baso name Output basis solution le name.
-brni name name is the lename of a variable branch order le to be read.
-brno name name is the lename of a variable branch order le to be written.
-d name val Assigns the value val to the parameter named name.
-dbgmem name Name of memory debug le. Write memory debug information to le name.
-f Complete license information is printed.
105
106 APPENDIX A. MOSEK COMMAND LINE TOOL REFERENCE
-h Prints out help information for MOSEK.
-inti name Input integer solution le name.
-into name Output integer solution le name.
-itri name Input interior point solution le name.
-itro name Output interior point solution le name.
-info name Infeasible subproblem output le name.
-infrepo name Feasibility reparation output le
-pari name Input parameter le name. Equivalent to -p.
-paro name Output parameter le name.
-L name name of the license le.
-l name name of the license le.
-max Forces MOSEK to maximize the objective.
-min Forces MOSEK to minimize the objective.
-n Ignore errors in subsequent paramter settings.
-p name New parameter settings are read from a le named name.
-q name Name of a optional log le.
-r If the option is present, the program returns -1 if an error ocurred otherwise 0.
-rout name If the option is present, the program writes the return code to le name.
-sen file Perform sensitivity analysis based on le.
-silent As little information as possible is send to the terminal.
-v The MOSEK version number is printed and no optimization is performed.
-w If this options is included, then MOSEK will wait for a license.
-= Lists the parameter database.
-? Same as the -h option.
A.3. THE PARAMETER FILE 107
A.3 The parameter le
Occasionally system or algorithmic parameters in MOSEK should be changed be the user. One way
of the changing parameters is to use a so-called parameter le which is a plain text le. It can for
example can have the format
BEGIN MOSEK
% This is a comment.
% The subsequent line tells MOSEK that an optimal
% basis should be computed by the interior-point optimizer.
MSK_IPAR_INTPNT_BASIS MSK_BI_ALWAYS
MSK_DPAR_INTPNT_TOL_PFEAS 1.0e-9
END MOSEK
Note that the le begins with an BEGIN MOSEK and is terminated with an END MOSEK, this is required.
Moreover, everything that appears after an % is considered to be a comment and is ignored. Similarly,
empty lines are ignored. The important lines are those which begins with a valid MOSEK parameter
name such as MSK IPAR INTPNT BASIS. Immediately after parameter name follows the new value for
the parameter. All the MOSEK parameter names are listed in Appendix H.
A.3.1 Using the parameter le
The parameter le can be given any name, but let us assume it has the name mosek.par. If MOSEK
should use the parameter settings in that le, then -p mosek.par should be on the command line
when MOSEK is invoked. An example of such a command line is
mosek -p mosek.par afiro.mps
108 APPENDIX A. MOSEK COMMAND LINE TOOL REFERENCE
Appendix B
The MPS le format
MOSEK supports the standard MPS format with some extensions. For a detailed description of the
MPS format the book by Nazareth [18] is a good reference.
B.1 The MPS le format
The version of the MPS format supported by MOSEK allows specication of an optimization problem
on the form
l
c
Ax + q(x) u
c
,
l
x
x u
x
,
x (,
x
J
integer,
(B.1)
where
x R
n
is the vector of decision variables.
A R
mn
l
c
R
m
u
c
R
m
l
x
R
n
u
x
R
n
q : R
n
R is a vector of quadratic functions. Hence,
q
i
(x) = 1/2x
T
Q
i
x
where it is assumed that
Q
i
= (Q
i
)
T
. (B.2)
Please note the explicit 1/2 in the quadratic term and that Q
i
is required to be symmetric.
( is a convex cone.
109
110 APPENDIX B. THE MPS FILE FORMAT
1, 2, . . . , n is an index set of the integer-constrained variables.
An MPS le with one row and one column can be illustrated like this:
* 1 2 3 4 5 6
*23456789012345678901234567890123456789012345678901234567890
NAME [name]
OBJSENSE
[objsense]
OBJNAME
[objname]
ROWS
? [cname1]
COLUMNS
[vname1] [cname1] [value1] [vname3] [value2]
RHS
[name] [cname1] [value1] [cname2] [value2]
RANGES
QSECTION [cname1]
[vname1] [vname2] [value1] [vname3] [value2]
BOUNDS
?? [name] [vname1] [value1]
CSECTION [kname1] [value1] [ktype]
[vname1]
ENDATA
Here the names in capitals are keywords of the MPS format and names in brackets are custom dened
names or values. A couple of notes on the structure:
Fields: All items surrounded by brackets appear in elds. The elds named valueN are numerical
values. Hence, they must have the format
[+|-]XXXXXXX.XXXXXX[[e|E][+|-]XXX]
where
X = [0|1|2|3|4|5|6|7|8|9].
Sections: The MPS le consists of several sections where the names in capitals indicate the beginning
of a new section. For example, COLUMNS denotes the beginning of the columns section.
Comments: Lines starting with an * are comment lines and are ignored by MOSEK.
Keys: The question marks represent keys to be specied later.
Extensions: The sections QSECTION and CSECTION are MOSEK specic extensions of the MPS format.
The standard MPS format is a xed format, i.e. everything in the MPS le must be within certain
xed positions. MOSEK also supports a free format. See Section B.5 for details.
B.1. THE MPS FILE FORMAT 111
B.1.1 An example
A concrete example of a MPS le is presented below:
NAME EXAMPLE
OBJSENSE
MIN
ROWS
N obj
L c1
L c2
L c3
L c4
COLUMNS
x1 obj -10.0 c1 0.7
x1 c2 0.5 c3 1.0
x1 c4 0.1
x2 obj -9.0 c1 1.0
x2 c2 0.8333333333 c3 0.66666667
x2 c4 0.25
RHS
rhs c1 630.0 c2 600.0
rhs c3 708.0 c4 135.0
ENDATA
Subsequently each individual section in the MPS format is discussed.
B.1.2 NAME
In this section a name ([name]) is assigned to the problem.
B.1.3 OBJSENSE (optional)
This is an optional section that can be used to specify the sense of the objective function. The OBJSENSE
section contains one line at most which can be one of the following
MIN
MINIMIZE
MAX
MAXIMIZE
It should be obvious what the implication is of each of these four lines.
B.1.4 OBJNAME (optional)
This is an optional section that can be used to specify the name of the row that is used as objective
function. The OBJNAME section contains one line at most which has the form
objname
objname should be a valid row name.
B.1.5 ROWS
A record in the ROWS section has the form
? [cname1]
where the requirements for the elds are as follows:
Field Starting Maximum Re- Description
position width quired
? 2 1 Yes Constraint key
[cname1] 5 8 Yes Constraint name
Hence, in this section each constraint is assigned an unique name denoted by [cname1]. Please note
that [cname1] starts in position 5 and the eld can be at most 8 characters wide. An initial key (?)
must be present to specify the type of the constraint. The key can have the values E, G, L, or N whith
ther following interpretation:
Constraint l
c
i
u
c
i
type
E nite l
c
i
G nite
L nite
N
In the MPS format an objective vector is not specied explicitly, but one of the constraints having the
key N will be used as the objective vector c. In general, if multiple N type constraints are specied,
then the rst will be used as the objective vector c.
B.1.6 COLUMNS
In this section the elements of A are specied using one or more records having the form
[vname1] [cname1] [value1] [cname2] [value2]
where the requirements for each eld are as follows:
[vname1] 5 8 Yes Variable name
[value1] 25 12 Yes Numerical value
[cname2] 40 8 No Constraint name
[value2] 50 12 No Numerical value
Hence, a record species one or two elements a
ij
of A using the principle that [vname1] and [cname1]
determines j and i respectively. Please note that [cname1] must be a constraint name specied in
the ROWS section. Finally, [value1] denotes the numerical value of a
ij
. Another optional element
is specied by [cname2], and [value2] for the variable specied by [vname1]. Some important
comments are:
All elements belonging to one variable must be grouped together.
Zero elements of A should not be specied.
At least one element for each variable should be specied.
B.1.7 RHS (optional)
A record in this section has the format
where the requirements for each eld are as follows:
[name] 5 8 Yes Name of the RHS vector
The interpretation of a record is that [name] is the name of the RHS vector to be specied. In general,
several vectors can be specied. [cname1] denotes a constraint name previously specied in the ROWS
section. Now, assume that this name has been assigned to the ith constraint and v
1
denotes the value
specied by [value1], then the interpretation of v
1
is:
Constraint l
c
i
u
c
i
type
E v
1
v
1
G v
1
L v
1
N
An optional second element is specied by [cname2] and [value2] and is interpreted in the same way.
Please note that it is not necessary to specify zero elements, because elements are assumed to be zero.
B.1.8 RANGES (optional)
A record in this section has the form
where the requirements for each elds are as follows:
[name] 5 8 Yes Name of the RANGE vector
The records in this section are used to modify the bound vectors for the constraints, i.e. the values
in l
c
and u
c
. A record has the following interpretation: [name] is the name of the RANGE vector anhd
[cname1] is a valid constraint name. Assume that [cname1] is assigned to the ith constraint and let
v
1
be the value specied by [value1], then a record has the interpretation:
Constraint Sign of v
1
l
c
i
u
c
i
type
E - u
c
i
+ v
1
E + l
c
i
+ v
1
G - or + l
c
i
+[v
1
[
L - or + u
c
i
[v
1
[
N
B.1.9 QSECTION (optional)
Within the QSECTION the label [cname1] must be a constraint name previously specied in the ROWS
section. The label [cname1] denotes the constraint to which the quadratic term belongs. A record in
the QSECTION has the form
[vname1] [vname2] [value1] [vname3] [value2]
where the requirements for each eld are:
[vname3] 40 8 No Variable name
A record species one or two elements in the lower triangular part of the Q
i
matrix where [cname1]
species the i. Hence, if the names [vname1] and [vname2] have been assigned to the kth and jth
variable, then Q
i
kj
is assigned the value given by [value1] An optional second element is specied in
the same way by the elds [vname1], [vname3], and [value2].
The example
minimize x
2
+ 0.5(2x
2
1
2x
1
x
3
+ 0.2x
2
2
+ 2x
2
3
)
subject to x
1
+ x
2
+ x
3
1,
x 0
has the following MPS le representation
NAME qoexp
ROWS
N obj
G c1
COLUMNS
x1 c1 1
x2 obj -1
x2 c1 1
x3 c1 1
RHS
rhs c1 1
QSECTION obj
x1 x1 2
x1 x3 -1
x2 x2 0.2
x3 x3 2
ENDATA
Regarding the QSECTIONs please note that:
Only one QSECTION is allowed for each constraint.
The QSECTIONs can appear in an arbitrary order after the COLUMNS section.
All variable names occurring in the QSECTION must already be specied in the COLUMNS section.
All entries specied in a QSECTION are assumed to belong to the lower triangular part of the
quadratic term of Q.
B.1.10 BOUNDS (optional)
In the BOUNDS section changes to the default bounds vectors l
x
and u
x
are specied. The default
bounds vectors are l
x
= 0 and u
x
= . Moreover, it is possible to specify several sets of bound
vectors. A record in this section has the form
?? [name] [vname1] [value1]
where the requirements for each eld are:
?? 2 2 Yes Bound key
[name] 5 8 Yes Name of the BOUNDS vector
[value1] 25 12 No Variable name
Hence, a record in the BOUNDS section has the following interpretation: [name] is the name of the
bound vector and [vname1] is the name of the variable which bounds are modied by the record. ??
and [value1] are used to modify the bound vectors according to the following table:
?? l
x
j
u
x
j
Made integer
(added to )
FR No
FX v
1
v
1
No
LO v
1
unchanged No
MI unchanged No
PL unchanged No
UP unchanged v
1
No
BV 0 1 Yes
LI v
1
| Yes
UI unchanged v
1
| Yes
v
1
is the value specied by [value1].
B.1.11 CSECTION (optional)
The purpose of the CSECTION is to specify the constraint
x (.
in (B.1).
It is assumed that ( satises the following requirements. Let
x
t
R
n
t
, t = 1, . . . , k
be vectors comprised of parts of the decision variables x so that each decision variable is a member of
exactly one vector x
t
, for example
x
1
=
_
_
x
1
x
4
x
7
_
_
and x
2
=
_
_
x
6
x
5
x
3
x
2
_
_
.
Next dene
( :=
_
x R
n
: x
t
(
t
, t = 1, . . . , k
_
where (
t
must have one of the following forms
R set:
(
t
= x R
n
t
.
Quadratic cone:
(
t
=
_
_
_
x R
n
t
: x
1

_
n
t
j=2
x
2
j
_
_
_
. (B.3)
(
t
=
_
_
_
x R
n
t
: 2x
1
x
2

n
t
j=3
x
2
j
, x
1
, x
2
0
_
_
_
. (B.4)
In general, only quadratic and rotated quadratic cones are specied in the MPS le whereas membership
of the R set is not. If a variable is not a member of any other cone then it is assumed to be a member
of an R cone.
Next, let us study an example. Assume that the quadratic cone
x
4

_
x
2
5
+ x
2
8
(B.5)
and the rotated quadratic cone
2x
3
x
7
x
2
1
+ x
2
8
, x
3
, x
7
0, (B.6)
should be specied in the MPS le. One CSECTION is required for each cone and they are specied as
follows:
* 1 2 3 4 5 6
*23456789012345678901234567890123456789012345678901234567890
CSECTION konea 0.0 QUAD
x4
x5
x8
CSECTION koneb 0.0 RQUAD
x7
x3
x1
x0
This rst CSECTION species the cone (B.5) which is given the name konea. This is a quadratic cone
which is specied by the keyword QUAD in the CSECTION header. The 0.0 value in the CSECTION header
is not used by the QUAD cone.
The second CSECTION species the rotated quadratic cone (B.6). Please note the keyword RQUAD in
the CSECTION which is used to specify that the cone is a rotated quadratic cone instead of a quadratic
cone. The 0.0 value in the CSECTION header is not used by the RQUAD cone.
In general, a CSECTION header has the format
CSECTION [kname1] [value1] [ktype]
where the requirement for each eld are as follows:
[kname1] 5 8 Yes Name of the cone
[value1] 15 12 No Cone parameter
[ktype] 25 Yes Type of the cone.
The possible cone type keys are:
Cone type key Members Interpretation.
QUAD 1 Quadratic cone i.e. (B.3).
RQUAD 2 Rotated quadratic cone i.e. (B.4).
Please note that a quadratic cone must have at least one member whereas a rotated quadratic cone
must have at least two members. A record in the CSECTION has the format
[vname1]
where the requirements for each eld are
[vname1] 2 8 Yes A valid variable name
The most important restriction with respect to the CSECTION is that a variable must occur in only
one CSECTION.
B.1.12 ENDATA
This keyword denotes the end of the MPS le.
B.2 Integer variables
Using special bound keys in the BOUNDS section it is possible to specify that some or all of the variables
should be integer-constrained i.e. be members of . However, an alternative method is available.
This method is available only for backward compability and we recommend that it is not used.
This method requires that markers are placed in the COLUMNS section as in the example:
COLUMNS
x1 obj -10.0 c1 0.7
x1 c2 0.5 c3 1.0
x1 c4 0.1
* Start of integer-constrained variables.
MARK000 MARKER INTORG
x2 obj -9.0 c1 1.0
x2 c2 0.8333333333 c3 0.66666667
x2 c4 0.25
x3 obj 1.0 c6 2.0
MARK001 MARKER INTEND
* End of integer-constrained variables.
Please note that special marker lines are used to indicate the start and the end of the integer
variables. Furthermore be aware of the following
IMPORTANT: All variables between the markers are assigned a default lower bound of 0 and
a default upper bound of 1. This may not be what is intended. If it is not intended, the
correct bounds should be dened in the BOUNDS section of the MPS formatted le.
MOSEK ignores eld 1, i.e. MARK0001 and MARK001, however, other optimization systems require
them.
Field 2, i.e. MARKER, must be specied including the single quotes. This implies that no row
can be assigned the name MARKER.
B.3. GENERAL LIMITATIONS 119
Field 3 is ignored and should be left blank.
Field 4, i.e. INTORG and INTEND, must be specied.
It is possible to specify several such integer marker sections within the COLUMNS section.
B.3 General limitations
An MPS le should be an ASCII le.
B.4 Interpretation of the MPS format
Several issues related to the MPS format are not well-dened by the industry standard. However,
MOSEK uses the following interpretation:
If a matrix element in the COLUMNS section is specied multiple times, then the multiple entries
are added together.
If a matrix element in a QSECTION section is specied multiple times, then the multiple entries
are added together.
B.5 The free MPS format
MOSEK supports a free format variation of the MPS format. The free format is similar to the MPS le
format but less restrictive, e.g. it allows longer names. However, it also presents two main limitations:
By default a line in the MPS le must not contain more than 1024 characters. However, by mod-
ifying the parameter MSK IPAR READ MPS WIDTH an arbitrary large line width will be accepted.
A name must not contain any blanks.
To use the free MPS format instead of the default MPS format the MOSEK parameter MSK IPAR READ MPS FORMAT
should be changed.
Appendix C
The LP le format
MOSEK supports the LP le format with some extensions i.e. MOSEK can read and write LP
formatted les.
C.1 A warning
The LP format is not a well-dened standard and hence dierent optimization packages may interpre-
tate a specic LP formatted le dierently.
C.2 The LP le format
The LP le format can specify problems on the form
minimize/maximize c
T
x +
1
2
q
o
(x)
subject to l
c
Ax +
1
2
q(x) u
c
,
l
x
x u
x
,
x
J
integer,
where
x R
n
is the vector of decision variables.
c R
n
is the linear term in the objective.
q
o
: R
n
R is the quadratic term in the objective where
q
o
(x) = x
T
Q
o
x
and it is assumed that
Q
o
= (Q
o
)
T
. (C.1)
A R
mn
l
c
R
m
121
122 APPENDIX C. THE LP FILE FORMAT
u
c
R
m
l
x
R
n
u
x
R
n
q : R
n
R is a vector of quadratic functions. Hence,
q
i
(x) = x
T
Q
i
x
where it is assumed that
Q
i
= (Q
i
)
T
. (C.2)
1, 2, . . . , n is an index set of the integer constrained variables.
C.2.1 The sections
An LP formatted le contains a number of sections specifying the objective, constraints, variable
bounds, and variable types. The section keywords may be any mix of upper and lower case letters.
C.2.1.1 The objective
The rst section beginning with one of the keywords
max
maximum
maximize
min
minimum
minimize
denes the objective sense and the objective function, i.e.
c
T
x +
1
2
x
T
Q
o
x.
The objective may be given a name by writing
myname:
before the expressions. If no name is given, then the objective is named obj.
The objective function contains linear and quadratic terms. The linear terms are written as in the
example
4 x1 + x2 - 0.1 x3
and so forth. The quadratic terms are written in square brackets ([ ]) and are either squared or
multiplied as in the examples
x1 ^ 2
and
C.2. THE LP FILE FORMAT 123
x1 * x2
There may be zero or more pairs of brackets containing quadratic expressions.
An example of an objective section is:
minimize
myobj: 4 x1 + x2 - 0.1 x3 + [ x1 ^ 2 + 2.1 x1 * x2 ]/2
Please note that the quadratic expressions are multiplied with
1
2
, so that the above expression means
minimize 4x
1
+ x
2
0.1 x
3
+
1
2
(x
2
1
+ 2.1 x
1
x
2
)
If the same variable occurs more than once in the linear part, the coecients are added, so that 4 x1
+ 2 x1 is equivalent to 6 x1. In the quadratic expressions x1 * x2 is equivalent to x2 * x1 and as
in the linear part , if the same variables multiplied or squared occur several times their coecients are
added.
C.2.1.2 The constraints
The second section beginning with one of the keywords
subj to
subject to
s.t.
st
denes the linear constraint matrix (A) and the quadratic matrices (Q
i
).
A constraint contains a name (optional), expressions adhering to the same rules as in the objective
and a bound:
subject to
con1: x1 + x2 + [ x3 ^ 2 ]/2 <= 5.1
The bound type (here <=) may be any of <, <=, =, >, >= (< and <= mean the same), and the bound
may be any number.
In the standard LP format it is not possible to dene more than one bound, but MOSEK supports
dening ranged constraints by using double-colon (::) instead of a single-colon (:) after the
constraint name, i.e.
5 x
1
+ x
2
5 (C.3)
may be written as
con:: -5 < x_1 + x_2 < 5
By default MOSEK writes ranged constraints this way.
If the les must adhere to the LP standard, ranged constraints must either be split into upper
bounded and lower bounded constraints or be written as en equality with a slack variable. For example
the expression (C.3) may be written as
x
1
+ x
2
sl
1
= 0, 5 sl
1
5.
C.2.1.3 Bounds
Bounds on the variables can be specied in the bound section beginning with one of the keywords
bound
bounds
The bounds section is optional but should, if present, follow the subject to section. All variables
listed in the bounds section must occur in either the objective or a constraint.
The default lower and upper bounds are 0 and +. A variable may be declared free with the key-
word free, which means that the lower bound is and the upper bound is +. Furthermore it may
be assigned a nite lower and upper bound. The bound denitions for a given variable may be written in
one or two lines, and bounds can be any number or (written as +inf/-inf/+infinity/-infinity)
as in the example
bounds
x1 free
x2 <= 5
0.1 <= x2
x3 = 42
2 <= x4 < +inf
C.2.1.4 Variable types
The nal two sections are optional and must begin with one of the keywords
bin
binaries
binary
and
gen
general
Under general all integer variables are listed, and under binary all binary (integer variables with
bounds 0 and 1) are listed:
general
x1 x2
binary
x3 x4
Again, all variables listed in the binary or general sections must occur in either the objective or a
constraint.
C.2.1.5 Terminating section
Finally, an LP formatted le must be terminated with the keyword
end
C.2.1.6 An example
A simple example of an LP le with two variables, four constraints and one integer variable is:
minimize
-10 x1 -9 x2
subject to
0.7 x1 + x2 <= 630
0.5 x1 + 0.833 x2 <= 600
x1 + 0.667 x2 <= 708
0.1 x1 + 0.025 x2 <= 135
bounds
10 <= x1
x1 <= +inf
20 <= x2 <= 500
general
x1
end
C.2.2 LP format peculiarities
C.2.2.1 Comments
Anything on a line after a is ignored and is treated as a comment.
C.2.2.2 Names
A name for an objective, a constraint or a variable may contain the letters a-z, A-Z, the digits 0-9 and
the characters
!"#$%&()/,.;?@_{}|~
The rst character in a name must not be a number, a period or the letter e or E. Keywords must
not be used as names.
It is strongly recommended not to use double quotes (") in names.
C.2.2.3 Variable bounds
Specifying several upper or lower bounds on one variable is possible but MOSEK uses only the tightest
bounds. If a variable is xed (with =), then it is considered the tightest bound.
C.2.2.4 MOSEK specic extensions to the LP format
Some optimization software packages employ a more strict denition of the LP format that the one
used by MOSEK. The limitations imposed by the strict LP format are the following:
Quadratic terms in the constraints are not allowed.
Names can be only 16 characters long.
Lines must not exceed 255 characters in length.
If an LP formatted le created by MOSEK should satises the strict denition, then the parameter
MSK_IPAR_WRITE_LP_STRICT_FORMAT
should be set; note, however, that some problems cannot be written correctly as a strict LP formatted
le. For instance, all names are truncated to 16 characters and hence they may loose their uniqueness
and change the problem.
To get around some of the inconveniences converting from other problem formats, MOSEK allows
lines to contain 1024 characters and names may have any length (shorter than the 1024 characters).
Internally in MOSEK names may contain any (printable) character, many of which cannot be used
in LP names. Setting the parameters
MSK_IPAR_READ_LP_QUOTED_NAMES
and
MSK_IPAR_WRITE_LP_QUOTED_NAMES
allows MOSEK to use quoted names. The rst parameter tells MOSEK to remove quotes from quoted
names e.g, "x1", when reading LP formatted les. The second parameter tells MOSEK to put quotes
around any semi-illegal name (names beginning with a number or a period) and fully illegal name
(containing illegal characters). As double quote is a legal character in the LP format, quoting semi-
illegal names makes them legal in the pure LP format as long as they are still shorter than 16 characters.
Fully illegal names are still illegal in a pure LP le.
C.2.3 The strict LP format
The LP format is not a formal standard and dierent vendors have slightly dierent interpretations of
the LP format. To make MOSEKs denition of the LP format more compatible whith the denitions
of other vendors use the paramter setting
MSK_IPAR_WRITE_LP_STRICT_FORMAT MSK_ON
This setting may lead to truncation of some names and hence to an invalid LP le. The simple solution
to this problem is to use the paramter setting
MSK_IPAR_WRITE_GENERIC_NAMES MSK_ON
which will cause all names to be renamed systematically in the output le.
C.2.4 Formatting of an LP le
A few parameters control the visual formatting of LP les written by MOSEK in order to make it
easier to read the les. These parameters are
MSK_IPAR_WRITE_LP_LINE_WIDTH
MSK_IPAR_WRITE_LP_TERMS_PER_LINE
The rst parameter sets the maximum number of characters on a single line. The default value is 80
corresponding roughly to the width of a standard text document.
The second parameter sets the maximum number of terms per line; a term means a sign, a coe-
cient, and a name (for example + 42 elephants). The default value is 0, meaning that there is no
maximum.
C.2.4.1 Speeding up le reading
If the input le should be read as fast as possible using the least amount of memory, then it is important
to tell MOSEK how many non-zeros, variables and constraints the problem contains. These values can
be set using the parameters
MSK_IPAR_READ_CON
MSK_IPAR_READ_VAR
MSK_IPAR_READ_ANZ
MSK_IPAR_READ_QNZ
C.2.4.2 Unnamed constraints
Reading and writing an LP le with MOSEK may change it supercially. If an LP le contains
unnamed constraints or objective these are given their generic names when the le is read (however
unnamed constraints in MOSEK are written without names).
Appendix D
The OPF format
The Optimization Problem Format (OPF) is an alternative to LP and MPS les for specifying opti-
mization problems. It is row-oriented, inspired by the CPLEX LP format.
Apart from containing objective, constraints, bounds etc. it may contain complete or partial
solutions, comments and extra information relevant for solving the problem. It is designed to be easily
read and modied by hand and to be forward compatible with possible future extensions.
D.1 Intended use
The OPF le format is meant to replace several other les:
The LP le format. Any problem that can be written as an LP le can be written as an OPF le
to; furthermore it naturally accommodates ranged constraints and variables as well as arbitrary
characters in names, xed expressions in the objective, empty constraints, and conic constraints.
Parameter les. It is possible to specify integer, double and string parameters along with the
problem (or in a separate OPF le).
Solution les. It is possible to store a full or a partial solution in an OPF le and later reload it.
D.2 The le format
The format uses tags to structure data. A simple example with the basic sections may look like this:
[comment]
This is a comment. You may write almost anything here...
[/comment]
# This is a single-line comment.
[objective min myobj]
x + 3 y + x^2 + 3 y^2 + z + 1
[/objective]
129
130 APPENDIX D. THE OPF FORMAT
[constraints]
[con con01] 4 <= x + y [/con]
[/constraints]
[bounds]
[b] -10 <= x,y <= 10 [/b]
[cone quad] x,y,z [/cone]
[/bounds]
A scope is opened by a tag of the form [tag] and closed by a tag of the form [/tag]. An opening tag
may accept a list of unnamed and named arguments, for examples
[tag value] tag with one unnamed argument [/tag]
[tag arg=value] tag with one named argument in quotes [/tag]
Unnamed arguments are identied by their order, while named arguments may appear in any order,
but never before an unnamed argument. The value can be a quoted, single-quoted or double-quoted
text string, i.e.
[tag value] single-quoted value [/tag]
[tag arg=value] single-quoted value [/tag]
[tag "value"] double-quoted value [/tag]
[tag arg="value"] double-quoted value [/tag]
D.2.1 Sections
The recognized tags are
[comment] A comment section. This can contain almost any text: Between single quotes () or
double quotes (") any text may appear. Outside quotes the markup characters ([ and ]) must
be prexed by backslashes. Both single and double quotes may appear alone or inside a pair of
quotes if it is prexed by a backslash.
[objective] The objective function: This accepts one or two parameters, where the rst one
(in the above example min) is either min or max (regardless of case) and denes the objective
sense, and the second one (above myobj), if present, is the objective name. The section may
contain linear and quadratic expressions.
If several objectives are specied, all but the last are ignored.
[constraints] This does not directly contain any data, but may contain the subsection con
dening a linear constraint.
[con] denes a single constraint; if an argument is present ([con NAME]) this is used as the name
of the constraint, otherwise it is given a null-name. The section contains a constraint denition
written as linear and quadratic expressions with a lower bound, an upper bound, with both or
with an equality. Examples:
D.2. THE FILE FORMAT 131
[constraints]
[con con1] 0 <= x + y [/con]
[con con2] 0 >= x + y [/con]
[con con3] 0 <= x + y <= 10 [/con]
[con con4] x + y = 10 [/con]
[/constraints]
Constraint names are unique. If a constraint is apecied which has the same name as a previously
dened constraint, the new constraint replaces the existing one.
[bounds] This does not directly contain any data, but may contain the subsections b (linear
bounds on variables) and cone (quadratic cone).
[b]. Bound denition on one or several variables separated by comma (,). An upper or
lower bound on a variable replaces any earlier dened bound on that variable. If only one
bound (upper or lower) is given only this bound is replaced. This means that upper and
lower bounds can be specied separately. So the OPF bound denition:
[b] x,y >= -10 [/b]
[b] x,y <= 10 [/b]
results in the bound
10 x, y 10. (D.1)
[cone]. Currently, the supported cones are the quadratic cone and the rotated quadratic
cone A conic constraint is dened as a set of variables which belongs to a single unique
cone.
A quadratic cone of n variables x
1
, . . . , x
n
denes a constraint of the form
x
2
1
>
n
i=2
x
2
i
.
A rotated quadratic cone of n variables x
1
, . . . , x
n
denes a constraint of the form
x
1
x
2
>
n
i=3
x
2
i
.
A [bounds]-section example:
[bounds]
[b] 0 <= x,y <= 10 [/b] # ranged bound
[b] 10 >= x,y >= 0 [/b] # ranged bound
[b] 0 <= x,y <= inf [/b] # using inf
[b] x,y free [/b] # free variables
# Let (x,y,z,w) belong to the cone K
[cone quad] x,y,z,w [/cone] # quadratic cone
[cone rquad] x,y,z,w [/cone] # rotated quadratic cone
[/bounds]
By default all variables are free.
[variables] This denes an ordering of variables as they should appear in the problem. This
is simply a space-separated list of variable names.
[integer] This contains a space-separated list of variables and denes the constraint that the
listed variables must be integer values.
[hints] This may contain only non-essential data; for example estimates of the number of
variables, constraints and non-zeros. Placed before all other sections containing data this may
reduce the time spent reading the le.
In the hints section, any subsection which is not recognized by MOSEK is simply ignored. In
this section a hint in a subsection is dened as follows:
[hint ITEM] value [/hint]
where ITEM may be replaced by numvar (number of variables), numcon (number of linear/quadratic
constraints), numanz (number if linear non-zeros in constraints) and numqnz (number of quadratic
non-zeros in constraints).
[solutions] This section can contain a number of full or partial solutions to a problem, each
inside a [solution]-section. The syntax is
[solution SOLTYPE status=STATUS]...[/solution]
where SOLTYPE is one of the strings
interior, a non-basic solution,
basic, a basic solution,
integer, an integer solution,
and STATUS is one of the strings
UNKNOWN,
OPTIMAL,
INTEGER OPTIMAL,
PRIM FEAS,
DUAL FEAS,
PRIM AND DUAL FEAS,
NEAR OPTIMAL,
NEAR PRIM FEAS,
NEAR DUAL FEAS,
NEAR PRIM AND DUAL FEAS,
PRIM INFEAS CER,
DUAL INFEAS CER,
D.2. THE FILE FORMAT 133
NEAR PRIM INFEAS CER,
NEAR DUAL INFEAS CER,
NEAR INTEGER OPTIMAL.
Most of these values are irrelevant for input solutions; when constructing a solution for simplex
hot-start or an initial solution for a mixed integer problem the safe setting is UNKNOWN.
A [solution]-section contains [con] and [var] sections. Each [con] and [var] section denes
solution values for a single variable or constraint, each value written as
KEYWORD=value
where KEYWORD denes a solution item and value denes its value. Allowed keywords are as
follows:
sk. The status of the item, where the value is one of the following strings:
LOW, the item is on its lower bound.
UPR, the item is on its upper bound.
FIX, it is a xed item.
BAS, the item is in the basis.
SUPBAS, the item is super basic.
UNK, the status is unknown.
INF, the item is outside its bounds (infeasible).
lvl Denes the level of the item.
sl Denes the level of the variable associated with its lower bound.
su Denes the level of the variable associated with its upper bound.
sn Denes the level of the variable associated with its cone.
y Denes the level of the corresponding dual variable (for constraints only).
A [var] section should always contain the items sk and lvl, and optionally sl, su and sn.
A [con] section should always contain sk and lvl, and optionally sl, su and y.
An example of a solution section
[solution basic status=UNKNOWN]
[var x0] sk=LOW lvl=5.0 [/var]
[var x1] sk=UPR lvl=10.0 [/var]
[var x2] sk=SUPBAS lvl=2.0 sl=1.5 su=0.0 [/var]
[con c0] sk=LOW lvl=3.0 y=0.0 [/con]
[con c0] sk=UPR lvl=0.0 y=5.0 [/con]
[/solution]
[vendor] This contains solver/vendor specic data. It accepts one argument, which is a vendor
ID for MOSEK the ID is simply mosek and the section contains the subsection parameters
dening solver parameters. When reading a vendor section, any unknown vendor can be safely
ignored. This is described later.
Comments using the # may appear anywhere in the le. Between the # and the following line-break
any text may be written, including markup characters.
D.2.2 Numbers
Numbers, when used for parameter values or coecients, are written in the usual way by the printf
function. That is, they may be prexed by a sign (+ or -) and may contain an integer part, decimal
part and an exponent. The decimal point is always . (a dot). Some examples are
1
1.0
.0
1.
1e10
1e+10
1e-10
Some invalid examples are
e10 # invalid, must contain either integer or decimal part
. # invalid
.e10 # invalid
More formally, the following standard regular expression describes numbers as used:
[+|-]?([0-9]+[.][0-9]*|[.][0-9]+)([eE][+|-]?[0-9]+)?
D.2.3 Names
Variable names, constraint names and objective name may contain arbitrary characters, which in some
cases must be enclosed by quotes (single or double) that in turn must be preceded by a backslash.
Unquoted names must begin with a letter (a-z or A-Z) and contain only the following characters: the
letters a-z and A-Z, the digits 0-9, braces ( and ) and underscore ( ).
Some examples of legal names:
an_unqouted_name
another_name{123}
single qouted name
"double qouted name"
"name with \"qoute\" in it"
"name with []s in it"
D.3 Parameters section
In the vendor section solver parameters are dened inside the parameters subsection. Each parameter
is written as
[p PARAMETER_NAME] value [/p]
D.4. WRITING OPF FILES FROM MOSEK 135
where PARAMETER NAME is replaced by a MOSEK parameter name, usually of the form MSK IPAR ...,
MSK DPAR ... or MSK SPAR ..., and the value is replaced by the value of that parameter; both integer
values and named values may be used. Some simple examples are:
[vendor mosek]
[parameters]
[p MSK_IPAR_OPF_MAX_TERMS_PER_LINE] 10 [/p]
[p MSK_IPAR_OPF_WRITE_PARAMETERS] MSK_ON [/p]
[p MSK_DPAR_DATA_TOL_BOUND_INF] 1.0e18 [/p]
[/parameters]
[/vendor]
D.4 Writing OPF les from MOSEK
To write an OPF le set the parameter MSK IPAR WRITE DATA FORMAT to MSK DATA FORMAT OP as this
ensures that OPF format is used. Then modify the following parameters to dene what the le should
contain:
MSK IPAR OPF WRITE HEADER, include a small header with comments.
MSK IPAR OPF WRITE HINTS, include hints about the size of the problem.
MSK IPAR OPF WRITE PROBLEM, include the problem itself objective, constraints and bounds.
MSK IPAR OPF WRITE SOLUTIONS, include solutions if they are dened. If this is o, no solutions
are included.
MSK IPAR OPF WRITE SOL BAS, include basic solution, if dened.
MSK IPAR OPF WRITE SOL ITG, include integer solution, if dened.
MSK IPAR OPF WRITE SOL ITR, include interior solution, if dened.
MSK IPAR OPF WRITE PARAMETERS, include all parameter settings.
D.5 Examples
This section contains a set of small examples written in OPF and describing how to formulate linear,
quadratic and conic problems.
D.5.1 Linear example lo1.opf
Consider the example:
minimize 10x
1
9x
2
,
subject to 7/10x
1
+ 1x
2
630,
1/2x
1
+ 5/6x
2
600,
1x
1
+ 2/3x
2
708,
1/10x
1
+ 1/4x
2
135,
x
1
, x
2
0.
(D.2)
In the OPF format the example is displayed as shown below:
[comment]
Example lo1.mps converted to OPF.
[/ comment]
[hints]
# Give a hint about the size of the different elements in the problem.
# These need only be estimates , but in this case they are exact.
[/ hints]
[variables]
# All variables that will appear in the problem
x1 x2
[/ variables]
[objective minimize obj ]
- 10 x1 - 9 x2
[/ objective]
[constraints]
[con c1 ] 0.7 x1 + x2 <= 630 [/con]
[con c2 ] 0.5 x1 + 0.8333333333 x2 <= 600 [/con]
[con c3 ] x1 + 0.66666667 x2 <= 708 [/con]
[con c4 ] 0.1 x1 + 0.25 x2 <= 135 [/con]
[/ constraints]
[bounds]
# By default all variables are free. The following line will
# change this to all variables being nonnegative.
[b] 0 <= * [/b]
[/ bounds]
D.5.2 Quadratic example qo1.opf
An example of a quadratic optimization problem is
minimize x
2
1
+ 0.1x
2
2
+ x
2
3
x
1
x
3
x
2
subject to 1 x
1
+ x
2
+ x
3
,
x 0.
(D.3)
This can be formulated in opf as shown below.
[comment]
Example qo1.mps converted to OPF.
[/ comment]
[hints]
[/ hints]
D.5. EXAMPLES 137
[variables]
x1 x2 x3
[/ variables]
# The quadratic terms are often multiplied by 1/2,
# but this is not required.
- x2 + 0.5 ( 2 x1 ^ 2 - 2 x3 * x1 + 0.2 x2 ^ 2 + 2 x3 ^ 2 )
[/ objective]
[constraints]
[con c1 ] 1 <= x1 + x2 + x3 [/con]
[/ constraints]
[bounds]
[b] 0 <= * [/b]
[/ bounds]
D.5.3 Conic quadratic example cqo1.opf
Consider the example:
minimize 1x
1
+ 2x
2
subject to 2x
3
+ 4x
4
= 5,
x
2
5
2x
1
x
3
,
x
2
6
2x
2
x
4
,
x
5
= 1,
x
6
= 1,
x 0.
(D.4)
Please note that the type of the cones is dened by the parameter to [cone ...]; the content of
the cone-section is the names of variables that belong to the cone.
[comment]
Example cqo1.mps converted to OPF.
[/ comment]
[hints]
[/ hints]
[variables]
x1 x2 x3 x4 x5 x6
[/ variables]
x1 + 2 x2
[/ objective]
[constraints]
[con c1 ] 2 x3 + 4 x4 = 5 [/con]
[/ constraints]
[bounds]
# We let all variables default to the positive orthant
[b] 0 <= * [/b]
# ... and change those that differ from the default.
[b] x5 ,x6 = 1 [/b]
# We define two rotated quadratic cones
# k1: 2 x1 * x3 >= x5^2
[cone rquad k1 ] x1, x3 , x5 [/cone]
# k2: 2 x2 * x4 >= x6^2
[cone rquad k2 ] x2, x4 , x6 [/cone]
[/ bounds]
D.5.4 Mixed integer example milo1.opf
Consider the mixed integer problem:
maximize x
0
+ 0.64x
1
subject to 50x
0
+ 31x
1
250,
3x
0
2x
1
4,
x
0
, x
1
0 and integer
(D.5)
This can be implemented in OPF with:
[comment]
Written by MOSEK version 5.0.0.7
Date 20-11-06
Time 14:42:24
[/ comment]
[hints]
[/ hints]
[variables disallow_new_variables]
x1 x2
[/ variables]
[objective maximize obj ]
x1 + 6.4e-1 x2
[/ objective]
[constraints]
[con c1 ] 5e+1 x1 + 3.1e+1 x2 <= 2.5e+2 [/con]
[con c2 ] -4 <= 3 x1 - 2 x2 [/con]
[/ constraints]
[bounds]
[b] 0 <= * [/b]
[/ bounds]
[integer]
D.5. EXAMPLES 139
x1 x2
[/ integer]
Appendix E
The XML (OSiL) format
MOSEK can write data in the standard OSiL xml format. For a denition of the OSiL format please see
http://www.optimizationservices.org/. Only linear constraints (possibly with integer variables)
are supported. By default output les with the extension .xml are written in the OSiL format.
The parameter MSK IPAR WRITE XML MODE controls if the linear coecients in the A matrix are
written in row or column order.
141
142 APPENDIX E. THE XML (OSIL) FORMAT
Appendix F
The solution le format
MOSEK provides one or two solution les depending on the problem type and the optimizer used.
If a problem is optimized using the interior-point optimizer and no basis identication is required,
then a le named probname.sol is provided. probname is the name of the problem and .sol is
the le extension. If the problem is optimized using the simplex optimizer or basis identication is
performed, then a le named probname.bas is created presenting the optimal basis solution. Finally,
if the problem contains integer constrained variables then a le named probname.int is created. It
contains the integer solution.
F.1 The basic and interior solution les
In general both the interior-point and the basis solution les have the format:
NAME : <problem name>
PROBLEM STATUS : <status of the problem>
SOLUTION STATUS : <status of the solution>
OBJECTIVE NAME : <name of the objective function>
PRIMAL OBJECTIVE : <primal objective value corresponding to the solution>
DUAL OBJECTIVE : <dual objective value corresponding to the solution>
CONSTRAINTS
? <name> ?? <a value> <a value> <a value> <a value> <a value>
VARIABLES
INDEX NAME AT ACTIVITY LOWER LIMIT UPPER LIMIT DUAL LOWER DUAL UPPER CONIC DUAL
? <name> ?? <a value> <a value> <a value> <a value> <a value> <a value>
In the example the elds ? and <> will be lled with problem and solution specic information. As
can be observed a solution report consists of three sections, i.e.
HEADER In this section, rst the name of the problem is listed and afterwards the problem and solution
statuses are shown. In this case the information shows that the problem is primal and dual
feasible and the solution is optimal. Next the primal and dual objective values are displayed.
CONSTRAINTS Subsequently in the constraint section the following information is listed for each con-
straint:
INDEX A sequential index assigned to the constraint by MOSEK.
NAME The name of the constraint assigned by the user.
AT The status of the constraint. In Table F.1 the possible values of the status keys and their
interpretation are shown.
143
144 APPENDIX F. THE SOLUTION FILE FORMAT
Status key Interpretation
UN Unknown status
BS Is basic
SB Is superbasic
LL Is at the lower limit (bound)
UL Is at the upper limit (bound)
EQ Lower limit is identical to upper limit
** Is infeasible i.e. the lower limit is
greater than the upper limit.
Table F.1: Status keys.
ACTIVITY Given the ith constraint on the form
l
c
i

n
j=1
a
ij
x
j
u
c
i
, (F.1)
then activity denote the quantity

n
j=1
a
ij
x
j
, where x
is the value for the x solution.

LOWER LIMIT Is the quantity l
c
i
(see (F.1)).
UPPER LIMIT Is the quantity u
c
i
(see (F.1)).
DUAL LOWER Is the dual multiplier corresponding to the lower limit on the constraint.
DUAL UPPER Is the dual multiplier corresponding to the upper limit on the constraint.
VARIABLES The last section of the solution report lists information for the variables. This information
has a similar interpretation as for the constraints. However, the column with the header [CONIC
DUAL] is only included for problems having one or more conic constraints. This column shows
the dual variables corresponding to the conic constraints.
F.2 The integer solution le
The integer solution is equivalent to the basic and interior solution les except that no dual information
is included.
Appendix G
The ORD le format
An ORD formatted le species in which order the mixed integer optimizer branches on variables. The
format of an ORD le is shown in Figure G.1. In the gure names in capitals are keywords of the ORD
format, whereas names in brackets are custom names or values. The ?? is an optional key specifying
the preferred branching direction. The possible keys are DN and UP which indicate that down or up
is the preferred branching direction respectively. The branching direction key is optional and is left
blank the mixed integer optimizer will decide whether to branch up or down.
* 1 2 3 4 5 6
*23456789012345678901234567890123456789012345678901234567890
NAME [name]
?? [vname1] [value1]
ENDATA
Figure G.1: The standard ORD format.
G.1 An example
A concrete example of a ORD le is presented below:
NAME EXAMPLE
DN x1 2
UP x2 1
x3 10
ENDATA
This implies that the priorities 2, 1, and 10 are assigned to variable x1, x2, and x3 respectively. The
higher the priority value assigned to a variable the earlier the mixed integer optimizer will branch on
that variable. The key DN implies that the mixed integer optimizer rst will branch down on variable
whereas the key UP implies that the mixed integer optimizer will rst branch up on a variable.
145
146 APPENDIX G. THE ORD FILE FORMAT
If no branch direction is specied for a variable then the mixed integer optimizer will automatically
choose the branching direction for that variable. Similarly, if no priority is assigned to a variable then
it is automatically assigned the priority of 0.
Appendix H
Parameters reference
Subsequently all parameters that are in MOSEK parameter database is presented. For each parameter
their name, purpose, type, default value etc. are presented.
H.1 Parameter groups
Parameters grouped by meaning and functionality.
H.1.1 Logging parameters.
MSK IPAR LOG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Controls the amount of log information.
MSK IPAR LOG BI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Controls the amount of output printed by the basis identication procedure. A higher level
implies that more information is logged.
MSK IPAR LOG BI FREQ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Controls the logging frequency.
MSK IPAR LOG CONCURRENT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
Controls amount of output printed by the concurrent optimizer.
MSK IPAR LOG CUT SECOND OPT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
Controls the reduction in the log levels for the second and any subsequent optimizations.
MSK IPAR LOG FACTOR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
If turned on, then the factor log lines are added to the log.
MSK IPAR LOG FEASREPAIR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Controls the amount of output printed when performing feasibility repair.
MSK IPAR LOG FILE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
If turned on, then some log info is printed when a le is written or read.
147
148 APPENDIX H. PARAMETERS REFERENCE
MSK IPAR LOG HEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
If turned on, then a header line is added to the log.
MSK IPAR LOG INFEAS ANA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Controls log level for the infeasibility analyzer.
MSK IPAR LOG INTPNT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Controls the amount of log information from the interior-point optimizers.
MSK IPAR LOG MIO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Controls the amount of log information from the mixed-integer optimizers.
MSK IPAR LOG MIO FREQ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227
The mixed-integer solver logging frequency.
MSK IPAR LOG NONCONVEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Controls amount of output printed by the nonconvex optimizer.
MSK IPAR LOG OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Controls the amount of general optimizer information that is logged.
MSK IPAR LOG ORDER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
If turned on, then factor lines are added to the log.
MSK IPAR LOG PARAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Controls the amount of information printed out about parameter changes.
MSK IPAR LOG PRESOLVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Controls amount of output printed by the presolve procedure. A higher level implies that more
information is logged.
MSK IPAR LOG RESPONSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Controls amount of output printed when response codes are reported. A higher level implies that
more information is logged.
MSK IPAR LOG SENSITIVITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Control logging in sensitivity analyzer.
MSK IPAR LOG SENSITIVITY OPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
MSK IPAR LOG SIM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Controls the amount of log information from the simplex optimizers.
MSK IPAR LOG SIM FREQ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230
Controls simplex logging frequency.
MSK IPAR LOG SIM NETWORK FREQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
Controls the network simplex logging frequency.
MSK IPAR LOG STORAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Controls the memory related log information.
H.1. PARAMETER GROUPS 149
H.1.2 Basis identication parameters.
MSK IPAR BI CLEAN OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Controls which simplex optimizer is used in the clean-up phase.
MSK IPAR BI IGNORE MAX ITER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Turns on basis identication in case the interior-point optimizer is terminated due to maximum
number of iterations.
MSK IPAR BI IGNORE NUM ERROR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Turns on basis identication in case the interior-point optimizer is terminated due to a numerical
problem.
MSK IPAR BI MAX ITERATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Maximum number of iterations after basis identication.
MSK IPAR INTPNT BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Controls whether basis identication is performed.
MSK IPAR LOG BI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
MSK DPAR SIM LU TOL REL PIV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Relative pivot tolerance employed when computing the LU factorization of the basis matrix.
H.1.3 The Interior-point method parameters.
Parameters dening the behavior of the interior-point method for linear, conic and convex problems.
problem.
MSK DPAR CHECK CONVEXITY REL TOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Convexity check tolerance.
MSK DPAR INTPNT CO TOL DFEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Dual feasibility tolerance used by the conic interior-point optimizer.
MSK DPAR INTPNT CO TOL INFEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Infeasibility tolerance for the conic solver.
MSK DPAR INTPNT CO TOL MU RED. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Optimality tolerance for the conic solver.
MSK DPAR INTPNT CO TOL NEAR REL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
MSK DPAR INTPNT CO TOL PFEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Primal feasibility tolerance used by the conic interior-point optimizer.
MSK DPAR INTPNT CO TOL REL GAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Relative gap termination tolerance used by the conic interior-point optimizer.
MSK IPAR INTPNT DIFF STEP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Controls whether dierent step sizes are allowed in the primal and dual space.
MSK IPAR INTPNT MAX ITERATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Controls the maximum number of iterations allowed in the interior-point optimizer.
MSK IPAR INTPNT MAX NUM COR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Maximum number of correction steps.
MSK IPAR INTPNT MAX NUM REFINEMENT STEPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Maximum number of steps to be used by the iterative search direction renement.
MSK DPAR INTPNT NL MERIT BAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Controls if the complementarity and infeasibility is converging to zero at about equal rates.
MSK DPAR INTPNT NL TOL DFEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Dual feasibility tolerance used when a nonlinear model is solved.
MSK DPAR INTPNT NL TOL MU RED. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Relative complementarity gap tolerance.
MSK DPAR INTPNT NL TOL NEAR REL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Nonlinear solver optimality tolerance parameter.
MSK DPAR INTPNT NL TOL PFEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Primal feasibility tolerance used when a nonlinear model is solved.
MSK DPAR INTPNT NL TOL REL GAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Relative gap termination tolerance for nonlinear problems.
MSK DPAR INTPNT NL TOL REL STEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Relative step size to the boundary for general nonlinear optimization problems.
MSK IPAR INTPNT OFF COL TRH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Controls the aggressiveness of the oending column detection.
MSK IPAR INTPNT ORDER METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Controls the ordering strategy.
MSK IPAR INTPNT REGULARIZATION USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Controls whether regularization is allowed.
MSK IPAR INTPNT SCALING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Controls how the problem is scaled before the interior-point optimizer is used.
MSK IPAR INTPNT SOLVE FORM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Controls whether the primal or the dual problem is solved.
MSK IPAR INTPNT STARTING POINT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Starting point used by the interior-point optimizer.
MSK DPAR INTPNT TOL DFEAS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Dual feasibility tolerance used for linear and quadratic optimization problems.
MSK DPAR INTPNT TOL DSAFE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Controls the interior-point dual starting point.
MSK DPAR INTPNT TOL INFEAS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Nonlinear solver infeasibility tolerance parameter.
MSK DPAR INTPNT TOL MU RED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
MSK DPAR INTPNT TOL PATH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
interior-point centering aggressiveness.
MSK DPAR INTPNT TOL PFEAS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Primal feasibility tolerance used for linear and quadratic optimization problems.
MSK DPAR INTPNT TOL PSAFE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Controls the interior-point primal starting point.
MSK DPAR INTPNT TOL REL GAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Relative gap termination tolerance.
MSK DPAR INTPNT TOL REL STEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Relative step size to the boundary for linear and quadratic optimization problems.
MSK DPAR INTPNT TOL STEP SIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
If the step size falls below the value of this parameter, then the interior-point optimizer assumes
that it is stalled. It it does not not make any progress.
MSK DPAR QCQO REFORMULATE REL DROP TOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
This parameter determines when columns are dropped in incomplete cholesky factorization doing
reformulation of quadratic problems.
MSK IPAR QO SEPARABLE REFORMULATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Determine if Quadratic programing problems should be reformulated to separable form.
H.1.4 Simplex optimizer parameters.
Parameters dening the behavior of the simplex optimizer for linear problems.
MSK DPAR BASIS REL TOL S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Maximum relative dual bound violation allowed in an optimal basic solution.
MSK IPAR BASIS SOLVE USE PLUS ONE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Controls the sign of the columns in the basis matrix corresponding to slack variables.
MSK DPAR BASIS TOL S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Maximum absolute dual bound violation in an optimal basic solution.
MSK DPAR BASIS TOL X. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Maximum absolute primal bound violation allowed in an optimal basic solution.
MSK IPAR LOG SIM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
MSK IPAR LOG SIM MINOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
Currently not in use.
MSK IPAR SENSITIVITY OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Controls which optimizer is used for optimal partition sensitivity analysis.
MSK IPAR SIM BASIS FACTOR USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Controls whether a (LU) factorization of the basis is used in a hot-start. Forcing a refactor-
ization sometimes improves the stability of the simplex optimizers, but in most cases there is a
performance penanlty.
MSK IPAR SIM DEGEN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Controls how aggressively degeneration is handled.
MSK IPAR SIM DUAL PHASEONE METHOD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
An exprimental feature.
MSK IPAR SIM EXPLOIT DUPVEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Controls if the simplex optimizers are allowed to exploit duplicated columns.
MSK IPAR SIM HOTSTART . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Controls the type of hot-start that the simplex optimizer perform.
MSK IPAR SIM INTEGER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
MSK IPAR SIM MAX ITERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Maximum number of iterations that can be used by a simplex optimizer.
MSK IPAR SIM MAX NUM SETBACKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Controls how many set-backs that are allowed within a simplex optimizer.
MSK IPAR SIM NETWORK DETECT METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Controls which type of detection method the network extraction should use.
MSK IPAR SIM NON SINGULAR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Controls if the simplex optimizer ensures a non-singular basis, if possible.
MSK IPAR SIM PRIMAL PHASEONE METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
MSK IPAR SIM REFORMULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Controls if the simplex optimizers are allowed to reformulate the problem.
MSK IPAR SIM SAVE LU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Controls if the LU factorization stored should be replaced with the LU factorization corresponding
to the initial basis.
MSK IPAR SIM SCALING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Controls how much eort is used in scaling the problem before a simplex optimizer is used.
MSK IPAR SIM SCALING METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Controls how the problem is scaled before a simplex optimizer is used.
MSK IPAR SIM SOLVE FORM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Controls whether the primal or the dual problem is solved by the primal-/dual- simplex optimizer.
MSK IPAR SIM STABILITY PRIORITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Controls how high priority the numerical stability should be given.
MSK IPAR SIM SWITCH OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Controls the simplex behavior.
MSK DPAR SIMPLEX ABS TOL PIV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Absolute pivot tolerance employed by the simplex optimizers.
H.1.5 Primal simplex optimizer parameters.
Parameters dening the behavior of the primal simplex optimizer for linear problems.
MSK IPAR SIM PRIMAL CRASH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Controls the simplex crash.
MSK IPAR SIM PRIMAL RESTRICT SELECTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Controls how aggressively restricted selection is used.
MSK IPAR SIM PRIMAL SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Controls the primal simplex strategy.
H.1.6 Dual simplex optimizer parameters.
Parameters dening the behavior of the dual simplex optimizer for linear problems.
MSK IPAR SIM DUAL CRASH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Controls whether crashing is performed in the dual simplex optimizer.
MSK IPAR SIM DUAL RESTRICT SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
MSK IPAR SIM DUAL SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Controls the dual simplex strategy.
H.1.7 Network simplex optimizer parameters.
Parameters dening the behavior of the network simplex optimizer for linear problems.
MSK IPAR SIM NETWORK DETECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Level of aggressiveness of network detection.
MSK IPAR SIM NETWORK DETECT HOTSTART . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Level of aggressiveness of network detection in a simplex hot-start.
MSK IPAR SIM REFACTOR FREQ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Controls the basis refactoring frequency.
H.1.8 Nonlinear convex method parameters.
Parameters dening the behavior of the interior-point method for nonlinear convex problems.
MSK IPAR CHECK CONVEXITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Specify the level of convexity check on quadratic problems
MSK IPAR LOG CHECK CONVEXITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Controls logging in convexity check on quadratic problems. Set to a positive value to turn logging
on.
If a quadratic coecient matrix is found to violate the requirement of PSD (NSD) then a list
of negative (positive) pivot elements is printed. The absolute value of the pivot elements is also
shown.
H.1.9 The conic interior-point method parameters.
Parameters dening the behavior of the interior-point method for conic problems.
H.1.10 The mixed-integer optimization parameters.
MSK IPAR LOG MIO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
MSK IPAR MIO BRANCH DIR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Controls whether the mixed-integer optimizer is branching up or down by default.
MSK IPAR MIO BRANCH PRIORITIES USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Controls whether branching priorities are used by the mixed-integer optimizer.
MSK IPAR MIO CONSTRUCT SOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Controls if an initial mixed integer solution should be constructed from the values of the integer
variables.
MSK IPAR MIO CONT SOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .232
Controls the meaning of interior-point and basic solutions in mixed integer problems.
MSK IPAR MIO CUT LEVEL ROOT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Controls the cut level employed by the mixed-integer optimizer at the root node.
MSK IPAR MIO CUT LEVEL TREE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Controls the cut level employed by the mixed-integer optimizer in the tree.
MSK DPAR MIO DISABLE TERM TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Certain termination criteria is disabled within the mixed-integer optimizer for period time spec-
ied by the parameter.
MSK IPAR MIO FEASPUMP LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Controls the feasibility pump heuristic which is used to construct a good initial feasible solution.
MSK IPAR MIO HEURISTIC LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
Controls the heuristic employed by the mixed-integer optimizer to locate an initial integer feasible
solution.
MSK DPAR MIO HEURISTIC TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Time limit for the mixed-integer heuristics.
MSK IPAR MIO HOTSTART . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
Controls whether the integer optimizer is hot-started.
MSK IPAR MIO KEEP BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
Controls whether the integer presolve keeps bases in memory.
MSK IPAR MIO MAX NUM BRANCHES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Maximum number of branches allowed during the branch and bound search.
MSK IPAR MIO MAX NUM RELAXS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Maximum number of relaxations in branch and bound search.
MSK IPAR MIO MAX NUM SOLUTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Controls how many feasible solutions the mixed-integer optimizer investigates.
MSK DPAR MIO MAX TIME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190
Time limit for the mixed-integer optimizer.
MSK DPAR MIO MAX TIME APRX OPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
MSK DPAR MIO NEAR TOL ABS GAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Relaxed absolute optimality tolerance employed by the mixed-integer optimizer.
MSK DPAR MIO NEAR TOL REL GAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
The mixed-integer optimizer is terminated when this tolerance is satised.
MSK IPAR MIO NODE OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Controls which optimizer is employed at the non-root nodes in the mixed-integer optimizer.
MSK IPAR MIO NODE SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Controls the node selection strategy employed by the mixed-integer optimizer.
MSK IPAR MIO OPTIMIZER MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
MSK IPAR MIO PRESOLVE AGGREGATE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Controls whether problem aggregation is performed in the mixed-integer presolve.
MSK IPAR MIO PRESOLVE PROBING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Controls whether probing is employed by the mixed-integer presolve.
MSK IPAR MIO PRESOLVE USE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Controls whether presolve is performed by the mixed-integer optimizer.
MSK DPAR MIO REL ADD CUT LIMITED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Controls cut generation for mixed-integer optimizer.
MSK DPAR MIO REL GAP CONST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
This value is used to compute the relative gap for the solution to an integer optimization problem.
MSK IPAR MIO ROOT OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Controls which optimizer is employed at the root node in the mixed-integer optimizer.
MSK IPAR MIO STRONG BRANCH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
The depth from the root in which strong branching is employed.
MSK DPAR MIO TOL ABS GAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Absolute optimality tolerance employed by the mixed-integer optimizer.
MSK DPAR MIO TOL ABS RELAX INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Integer constraint tolerance.
MSK DPAR MIO TOL FEAS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
Feasibility tolerance for mixed integer solver. Any solution with maximum infeasibility below
this value will be considered feasible.
MSK DPAR MIO TOL REL GAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Relative optimality tolerance employed by the mixed-integer optimizer.
MSK DPAR MIO TOL REL RELAX INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
MSK DPAR MIO TOL X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Absolute solution tolerance used in mixed-integer optimizer.
H.1.11 Presolve parameters.
MSK IPAR PRESOLVE ELIM FILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Maximum amount of ll-in in the elimination phase.
MSK IPAR PRESOLVE ELIMINATOR MAX NUM TRIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Control the maximum number of times the eliminator is tried.
MSK IPAR PRESOLVE ELIMINATOR USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Controls whether free or implied free variables are eliminated from the problem.
MSK IPAR PRESOLVE LEVEL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Currently not used.
MSK IPAR PRESOLVE LINDEP USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Controls whether the linear constraints are checked for linear dependencies.
MSK IPAR PRESOLVE LINDEP WORK LIM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Controls linear dependency check in presolve.
MSK DPAR PRESOLVE TOL AIJ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Absolute zero tolerance employed for constraint coecients in the presolve.
MSK DPAR PRESOLVE TOL LIN DEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Controls when a constraint is determined to be linearly dependent.
MSK DPAR PRESOLVE TOL S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Absolute zero tolerance employed for slack variables in the presolve.
MSK DPAR PRESOLVE TOL X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Absolute zero tolerance employed for variables in the presolve.
MSK IPAR PRESOLVE USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Controls whether the presolve is applied to a problem before it is optimized.
H.1.12 Termination criterion parameters.
Parameters which dene termination and optimality criteria and related information.
MSK DPAR LOWER OBJ CUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Objective bound.
MSK DPAR LOWER OBJ CUT FINITE TRH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Objective bound.
MSK DPAR OPTIMIZER MAX TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Solver time limit.
MSK DPAR UPPER OBJ CUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Objective bound.
MSK DPAR UPPER OBJ CUT FINITE TRH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Objective bound.
H.1.13 Progress call-back parameters.
MSK DPAR CALLBACK FREQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Controls progress call-back frequency.
MSK IPAR SOLUTION CALLBACK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Indicates whether solution call-backs will be performed during the optimization.
H.1.14 Non-convex solver parameters.
MSK IPAR NONCONVEX MAX ITERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Maximum number of iterations that can be used by the nonconvex optimizer.
MSK DPAR NONCONVEX TOL FEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Feasibility tolerance used by the nonconvex optimizer.
MSK DPAR NONCONVEX TOL OPT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Optimality tolerance used by the nonconvex optimizer.
H.1.15 Feasibility repair parameters.
MSK DPAR FEASREPAIR TOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Tolerance for constraint enforcing upper bound on sum of weighted violations in feasibility repair.
H.1.16 Optimization system parameters.
Parameters dening the overall solver system environment. This includes system and platform related
information and behavior.
MSK IPAR AUTO UPDATE SOL INFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Controls whether the solution information items are automatically updated after an optimization
is performed.
MSK IPAR CACHE LICENSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Control license caching.
MSK IPAR CACHE SIZE L1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Species the size of the level 1 cache of the processor.
MSK IPAR CPU TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Species the CPU type.
MSK IPAR INTPNT NUM THREADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Controls the number of threads employed by the interior-point optimizer. If set to a positive
number MOSEK will use this number of threads. If zero the number of threads used will equal
the number of cores detected on the machine.
MSK IPAR LICENSE CACHE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Setting this parameter no longer has any eect.
MSK IPAR LICENSE CHECK TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Controls the license manager client behavior.
MSK IPAR LICENSE WAIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Controls if MOSEK should queue for a license if none is available.
MSK IPAR TIMING LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Controls the a amount of timing performed inside MOSEK.
H.1.17 Output information parameters.
MSK IPAR INFEAS REPORT LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Controls the contents of the infeasibility report.
MSK IPAR LICENSE SUPPRESS EXPIRE WRNS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Controls license manager client behavior.
MSK IPAR LOG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
MSK IPAR LOG BI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
MSK IPAR LOG FILE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
MSK IPAR LOG HEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
MSK IPAR LOG MIO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
MSK IPAR LOG SIM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
MSK IPAR MAX NUM WARNINGS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Waning level. A higher value results in more warnings.
MSK IPAR WARNING LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Warning level.
H.1.18 Extra information about the optimization problem.
MSK IPAR OBJECTIVE SENSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
If the objective sense for the task is undened, then the value of this parameter is used as the
default objective sense.
H.1.19 Overall solver parameters.
MSK IPAR CONCURRENT NUM OPTIMIZERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
The maximum number of simultaneous optimizations that will be started by the concurrent
optimizer.
MSK IPAR CONCURRENT PRIORITY DUAL SIMPLEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Priority of the dual simplex algorithm when selecting solvers for concurrent optimization.
MSK IPAR CONCURRENT PRIORITY FREE SIMPLEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Priority of the free simplex optimizer when selecting solvers for concurrent optimization.
MSK IPAR CONCURRENT PRIORITY INTPNT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Priority of the interior-point algorithm when selecting solvers for concurrent optimization.
MSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Priority of the primal simplex algorithm when selecting solvers for concurrent optimization.
MSK IPAR DATA CHECK. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Enable data checking for debug purposes.
MSK IPAR FEASREPAIR OPTIMIZE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Controls which type of feasibility analysis is to be performed.
MSK IPAR INFEAS PREFER PRIMAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Controls which certicate is used if both primal- and dual- certicate of infeasibility is available.
MSK IPAR MIO LOCAL BRANCH NUMBER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Controls the size of the local search space when doing local branching.
MSK IPAR MIO MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Turns on/o the mixed-integer mode.
MSK IPAR OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Controls which optimizer is used to optimize the task.
Currently not used.
MSK IPAR SENSITIVITY ALL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Controls sensitivity report behavior.
MSK IPAR SENSITIVITY TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Controls which type of sensitivity analysis is to be performed.
H.1.20 Behavior of the optimization task.
Parameters dening the behavior of an optimization task when loading data.
MSK IPAR ALLOC ADD QNZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Controls how the quadratic matrixes are extended.
MSK SPAR FEASREPAIR NAME PREFIX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Feasibility repair name prex.
MSK SPAR FEASREPAIR NAME SEPARATOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Feasibility repair name separator.
MSK SPAR FEASREPAIR NAME WSUMVIOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Feasibility repair name violation name.
MSK IPAR READ ADD ANZ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .245
Controls how the constraint matrix is extended.
MSK IPAR READ ADD CON. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246
Additional number of constraints that is made room for in the problem.
MSK IPAR READ ADD CONE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Additional number of conic constraints that is made room for in the problem.
MSK IPAR READ ADD QNZ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246
MSK IPAR READ ADD VAR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246
Additional number of variables that is made room for in the problem.
MSK IPAR READ ANZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Controls the expected number of constraint non-zeros.
MSK IPAR READ CON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Controls the expected number of constraints.
MSK IPAR READ CONE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Controls the expected number of conic constraints.
MSK IPAR READ QNZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Controls the expected number of quadratic non-zeros.
MSK IPAR READ TASK IGNORE PARAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Controls what information is used from the task les.
MSK IPAR READ VAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Controls the expected number of variables.
MSK IPAR WRITE TASK INC SOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Controls whether the solutions are stored in the task le too.
H.1.21 Data input/output parameters.
Parameters dening the behavior of data readers and writers.
MSK SPAR BAS SOL FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Name of the bas solution le.
MSK SPAR DATA FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Data are read and written to this le.
MSK SPAR DEBUG FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
MOSEK debug le.
MSK IPAR INFEAS REPORT AUTO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Turns the feasibility report on or o.
MSK SPAR INT SOL FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Name of the int solution le.
MSK SPAR ITR SOL FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Name of the itr solution le.
MSK IPAR LOG FILE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
MSK IPAR LP WRITE IGNORE INCOMPATIBLE ITEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Controls the result of writing a problem containing incompatible items to an LP le.
MSK IPAR OPF MAX TERMS PER LINE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
The maximum number of terms (linear and quadratic) per line when an OPF le is written.
MSK IPAR OPF WRITE HEADER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Write a text header with date and MOSEK version in an OPF le.
MSK IPAR OPF WRITE HINTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Write a hint section with problem dimensions in the beginning of an OPF le.
MSK IPAR OPF WRITE PARAMETERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Write a parameter section in an OPF le.
MSK IPAR OPF WRITE PROBLEM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Write objective, constraints, bounds etc. to an OPF le.
MSK IPAR OPF WRITE SOL BAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Controls what is written to the OPF les.
MSK IPAR OPF WRITE SOL ITG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
MSK IPAR OPF WRITE SOL ITR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
MSK IPAR OPF WRITE SOLUTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Enable inclusion of solutions in the OPF les.
MSK SPAR PARAM COMMENT SIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Solution le comment character.
MSK IPAR PARAM READ CASE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
If turned on, then names in the parameter le are case sensitive.
MSK SPAR PARAM READ FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Modications to the parameter database is read from this le.
MSK IPAR PARAM READ IGN ERROR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
If turned on, then errors in paramter settings is ignored.
MSK SPAR PARAM WRITE FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
The parameter database is written to this le.
MSK IPAR READ ANZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
MSK IPAR READ CON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
MSK IPAR READ DATA COMPRESSED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Controls the input le decompression.
MSK IPAR READ DATA FORMAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Format of the data le to be read.
MSK IPAR READ KEEP FREE CON. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Controls whether the free constraints are included in the problem.
MSK IPAR READ LP DROP NEW VARS IN BOU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Controls how the LP les are interpreted.
MSK IPAR READ LP QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
If a name is in quotes when reading an LP le, the quotes will be removed.
MSK SPAR READ MPS BOU NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Name of the BOUNDS vector used. An empty name means that the rst BOUNDS vector is
used.
MSK IPAR READ MPS FORMAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Controls how strictly the MPS le reader interprets the MPS format.
MSK IPAR READ MPS KEEP INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Controls if integer constraints are read.
MSK SPAR READ MPS OBJ NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Objective name in the MPS le.
MSK IPAR READ MPS OBJ SENSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Controls the MPS format extensions.
MSK IPAR READ MPS QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
MSK SPAR READ MPS RAN NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Name of the RANGE vector used. An empty name means that the rst RANGE vector is used.
MSK IPAR READ MPS RELAX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Controls the meaning of integer constraints.
MSK SPAR READ MPS RHS NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Name of the RHS used. An empty name means that the rst RHS vector is used.
MSK IPAR READ MPS WIDTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Controls the maximal number of characters allowed in one line of the MPS le.
MSK IPAR READ Q MODE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Controls how the Q matrices are read from the MPS le.
MSK IPAR READ QNZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
MSK IPAR READ VAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
MSK SPAR SENSITIVITY FILE NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Sensitivity report le name.
MSK SPAR SENSITIVITY RES FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Name of the sensitivity report output le.
MSK SPAR SOL FILTER XC LOW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Solution le lter.
MSK SPAR SOL FILTER XC UPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Solution le lter.
MSK SPAR SOL FILTER XX LOW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Solution le lter.
MSK SPAR SOL FILTER XX UPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Solution le lter.
MSK IPAR SOL QUOTED NAMES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Controls the solution le format.
MSK IPAR SOL READ NAME WIDTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Controls the input solution le format.
MSK IPAR SOL READ WIDTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
MSK SPAR STAT FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Statistics le name.
MSK SPAR STAT KEY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Key used when writing the summary le.
MSK SPAR STAT NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Name used when writing the statistics le.
MSK IPAR WRITE BAS CONSTRAINTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Controls the basic solution le format.
MSK IPAR WRITE BAS HEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
MSK IPAR WRITE BAS VARIABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
MSK IPAR WRITE DATA COMPRESSED. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Controls output le compression.
MSK IPAR WRITE DATA FORMAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Controls the output le format.
MSK IPAR WRITE DATA PARAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Controls output le data.
MSK IPAR WRITE FREE CON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Controls the output le data.
MSK IPAR WRITE GENERIC NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
MSK IPAR WRITE GENERIC NAMES IO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Index origin used in generic names.
MSK IPAR WRITE INT CONSTRAINTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Controls the integer solution le format.
MSK IPAR WRITE INT HEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
MSK IPAR WRITE INT VARIABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
MSK SPAR WRITE LP GEN VAR NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Added variable names in the LP les.
MSK IPAR WRITE LP LINE WIDTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Controls the LP output le format.
MSK IPAR WRITE LP QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Controls LP output le format.
MSK IPAR WRITE LP STRICT FORMAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Controls whether LP output les satisfy the LP format strictly.
MSK IPAR WRITE LP TERMS PER LINE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
MSK IPAR WRITE MPS INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
MSK IPAR WRITE MPS OBJ SENSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
MSK IPAR WRITE MPS QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
MSK IPAR WRITE MPS STRICT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Controls the output MPS le format.
MSK IPAR WRITE PRECISION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Controls data precision employed in when writing an MPS le.
MSK IPAR WRITE SOL CONSTRAINTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
MSK IPAR WRITE SOL HEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Controls solution le format.
MSK IPAR WRITE SOL VARIABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
MSK IPAR WRITE XML MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Controls if linear coecients should be written by row or column when writing in the XML le
format.
H.1.22 Analysis parameters.
Parameters controling the behaviour of the problem and solution analyzers.
MSK IPAR ANA SOL BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Controls whether the basis matrix is analyzed in solaution analyzer.
MSK DPAR ANA SOL INFEAS TOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
If a constraint violates its bound with an amount larger than this value, the constraint name,
index and violation will be printed by the solution analyzer.
MSK IPAR ANA SOL PRINT VIOLATED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Controls whether a list of violated constraints is printed.
H.1.23 Solution input/output parameters.
Parameters dening the behavior of solution reader and writer.
MSK IPAR SOL FILTER KEEP BASIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
MSK IPAR SOL FILTER KEEP RANGED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Control the contents of the solution les.
Solution le lter.
Solution le lter.
Solution le lter.
Solution le lter.
H.1.24 Infeasibility report parameters.
MSK IPAR INFEAS GENERIC NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
H.1.25 License manager parameters.
MSK IPAR LICENSE ALLOW OVERUSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Controls if license overuse is allowed when caching licenses
MSK IPAR LICENSE DEBUG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Controls the license manager client debugging behavior.
MSK IPAR LICENSE PAUSE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
H.1.26 Data check parameters.
These parameters denes data checking settings and problem data tolerances, i.e. which values are
rounded to 0 or innity, and which values are large or small enough to produce a warning.
MSK IPAR CHECK TASK DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
If this feature is turned on, then the task data is checked for bad values i.e. NaNs. before an
MSK DPAR DATA TOL AIJ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180
Data tolerance threshold.
MSK DPAR DATA TOL AIJ HUGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
MSK DPAR DATA TOL AIJ LARGE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
MSK DPAR DATA TOL BOUND INF. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL BOUND WRN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL C HUGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL CJ LARGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL QIJ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182
MSK DPAR DATA TOL X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
on.
shown.
H.1.27 Debugging parameters.
These parameters denes that can be used when debugging a problem.
MSK IPAR AUTO SORT A BEFORE OPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Controls whether the elements in each column of A are sorted before an optimization is performed.
H.2 Double parameters
MSK DPAR ANA SOL INFEAS TOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
If a constraint violates its bound with an amount larger than this value, the constraint name,
index and violation will be printed by the solution analyzer.
MSK DPAR CALLBACK FREQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Controls progress call-back frequency.
H.2. DOUBLE PARAMETERS 175
MSK DPAR CHECK CONVEXITY REL TOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Convexity check tolerance.
MSK DPAR DATA TOL AIJ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180
MSK DPAR DATA TOL AIJ HUGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
MSK DPAR DATA TOL AIJ LARGE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
MSK DPAR DATA TOL BOUND INF. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL BOUND WRN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL C HUGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL CJ LARGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
MSK DPAR DATA TOL QIJ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182
MSK DPAR DATA TOL X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
MSK DPAR FEASREPAIR TOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Tolerance for constraint enforcing upper bound on sum of weighted violations in feasibility repair.
MSK DPAR INTPNT TOL DSAFE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Controls the interior-point dual starting point.
MSK DPAR INTPNT TOL PATH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
interior-point centering aggressiveness.
MSK DPAR INTPNT TOL PSAFE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Controls the interior-point primal starting point.
MSK DPAR INTPNT TOL REL STEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
MSK DPAR INTPNT TOL STEP SIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
If the step size falls below the value of this parameter, then the interior-point optimizer assumes
that it is stalled. It it does not not make any progress.
MSK DPAR LOWER OBJ CUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Objective bound.
MSK DPAR LOWER OBJ CUT FINITE TRH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Objective bound.
MSK DPAR MIO HEURISTIC TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Time limit for the mixed-integer heuristics.
MSK DPAR MIO MAX TIME APRX OPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
MSK DPAR MIO NEAR TOL ABS GAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Relaxed absolute optimality tolerance employed by the mixed-integer optimizer.
MSK DPAR MIO REL ADD CUT LIMITED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Controls cut generation for mixed-integer optimizer.
MSK DPAR MIO TOL ABS GAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
MSK DPAR MIO TOL ABS RELAX INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
MSK DPAR MIO TOL FEAS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
Feasibility tolerance for mixed integer solver. Any solution with maximum infeasibility below
this value will be considered feasible.
MSK DPAR MIO TOL REL RELAX INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
MSK DPAR MIO TOL X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
MSK DPAR NONCONVEX TOL FEAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
MSK DPAR NONCONVEX TOL OPT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
MSK DPAR OPTIMIZER MAX TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Solver time limit.
MSK DPAR PRESOLVE TOL AIJ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Absolute zero tolerance employed for constraint coecients in the presolve.
MSK DPAR PRESOLVE TOL LIN DEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
MSK DPAR PRESOLVE TOL S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Absolute zero tolerance employed for slack variables in the presolve.
MSK DPAR PRESOLVE TOL X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Absolute zero tolerance employed for variables in the presolve.
MSK DPAR QCQO REFORMULATE REL DROP TOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
This parameter determines when columns are dropped in incomplete cholesky factorization doing
reformulation of quadratic problems.
MSK DPAR SIMPLEX ABS TOL PIV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
MSK DPAR UPPER OBJ CUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Objective bound.
MSK DPAR UPPER OBJ CUT FINITE TRH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Objective bound.
ana sol infeas tol
Corresponding constant:
MSK DPAR ANA SOL INFEAS TOL
Description:
If a constraint violates its bound with an amount larger than this value, the constraint
name, index and violation will be printed by the solution analyzer.
Possible Values:
Any number between 0.0 and +inf.
Default value:
+1e-8
basis rel tol s
MSK DPAR BASIS REL TOL S
Description:
Possible Values:
Default value:
1.0e-12
basis tol s
MSK DPAR BASIS TOL S
Description:
Possible Values:
Any number between 1.0e-9 and +inf.
Default value:
1.0e-6
basis tol x
MSK DPAR BASIS TOL X
Description:
Possible Values:
Default value:
1.0e-6
callback freq
MSK DPAR CALLBACK FREQ
Description:
Controls the time between calls to the progress call-back function. Hence, if the value of
this parameter is for example 10, then the call-back is called approximately each 10 seconds.
A negative value is equivalent to innity.
In general frequent call-backs may hurt the performance.
Possible Values:
Any number between -inf and +inf.
Default value:
-1.0
check convexity rel tol
MSK DPAR CHECK CONVEXITY REL TOL
Description:
This parameter controls when the full convexity check declares a problem to be non-convex.
Increasing this tolerance relaxes the criteria for declaring the problem non-convex.
A problem is declared non-convex if negative (positive) pivot elements are detected in the
cholesky factor of a matrix which is required to be PSD (NSD). This parameter controles
how much this non-negativity requirement may be violated.
If d
i
is the pivot element for column i, then the matrix Q is considered to not be PSD if:
d
i
[Q
ii
[ check convexity rel tol
Possible Values:
Any number between 0 and +inf.
Default value:
1e-10
data tol aij
MSK DPAR DATA TOL AIJ
Description:
Absolute zero tolerance for elements in A. If any value A
ij
is smaller than this parameter
in absolute terms MOSEK will treat the values as zero and generate a warning.
Possible Values:
Any number between 1.0e-16 and 1.0e-6.
Default value:
1.0e-12
data tol aij huge
MSK DPAR DATA TOL AIJ HUGE
Description:
An element in A which is larger than this value in absolute size causes an error.
Possible Values:
Default value:
1.0e20
data tol aij large
MSK DPAR DATA TOL AIJ LARGE
Description:
An element in A which is larger than this value in absolute size causes a warning message
to be printed.
Possible Values:
Default value:
1.0e10
data tol bound inf
MSK DPAR DATA TOL BOUND INF
Description:
Any bound which in absolute value is greater than this parameter is considered innite.
Possible Values:
Default value:
1.0e16
data tol bound wrn
MSK DPAR DATA TOL BOUND WRN
Description:
If a bound value is larger than this value in absolute size, then a warning message is issued.
Possible Values:
Default value:
1.0e8
data tol c huge
MSK DPAR DATA TOL C HUGE
Description:
An element in c which is larger than the value of this parameter in absolute terms is
considered to be huge and generates an error.
Possible Values:
Default value:
1.0e16
data tol cj large
MSK DPAR DATA TOL CJ LARGE
Description:
An element in c which is larger than this value in absolute terms causes a warning message
to be printed.
Possible Values:
Default value:
1.0e8
data tol qij
MSK DPAR DATA TOL QIJ
Description:
Absolute zero tolerance for elements in Q matrices.
Possible Values:
Default value:
1.0e-16
data tol x
MSK DPAR DATA TOL X
Description:
Zero tolerance for constraints and variables i.e. if the distance between the lower and upper
bound is less than this value, then the lower and lower bound is considered identical.
Possible Values:
Default value:
1.0e-8
feasrepair tol
MSK DPAR FEASREPAIR TOL
Description:
Tolerance for constraint enforcing upper bound on sum of weighted violations in feasibility
repair.
Possible Values:
Any number between 1.0e-16 and 1.0e+16.
Default value:
1.0e-10
intpnt co tol dfeas
MSK DPAR INTPNT CO TOL DFEAS
Description:
Possible Values:
Any number between 0.0 and 1.0.
Default value:
1.0e-8
See also:
MSK DPAR INTPNT CO TOL NEAR REL Optimality tolerance for the conic solver.
intpnt co tol infeas
MSK DPAR INTPNT CO TOL INFEAS
Description:
Controls when the conic interior-point optimizer declares the model primal or dual infeasible.
A small number means the optimizer gets more conservative about declaring the model
infeasible.
Possible Values:
Default value:
1.0e-8
intpnt co tol mu red
MSK DPAR INTPNT CO TOL MU RED
Description:
Relative complementarity gap tolerance feasibility tolerance used by the conic interior-point
optimizer.
Possible Values:
Default value:
1.0e-8
intpnt co tol near rel
MSK DPAR INTPNT CO TOL NEAR REL
Description:
If MOSEK cannot compute a solution that has the prescribed accuracy, then it will multiply
the termination tolerances with value of this parameter. If the solution then satises the
termination criteria, then the solution is denoted near optimal, near feasible and so forth.
Possible Values:
Default value:
100
intpnt co tol pfeas
MSK DPAR INTPNT CO TOL PFEAS
Description:
Possible Values:
Default value:
1.0e-8
See also:
intpnt co tol rel gap
MSK DPAR INTPNT CO TOL REL GAP
Description:
Possible Values:
Default value:
1.0e-8
See also:
intpnt nl merit bal
MSK DPAR INTPNT NL MERIT BAL
Description:
Possible Values:
Default value:
1.0e-4
intpnt nl tol dfeas
MSK DPAR INTPNT NL TOL DFEAS
Description:
Possible Values:
Default value:
1.0e-8
intpnt nl tol mu red
MSK DPAR INTPNT NL TOL MU RED
Description:
Possible Values:
Default value:
1.0e-12
intpnt nl tol near rel
MSK DPAR INTPNT NL TOL NEAR REL
Description:
If the MOSEK nonlinear interior-point optimizer cannot compute a solution that has the
prescribed accuracy, then it will multiply the termination tolerances with value of this
parameter. If the solution then satises the termination criteria, then the solution is denoted
near optimal, near feasible and so forth.
Possible Values:
Default value:
1000.0
intpnt nl tol pfeas
MSK DPAR INTPNT NL TOL PFEAS
Description:
Possible Values:
Default value:
1.0e-8
intpnt nl tol rel gap
MSK DPAR INTPNT NL TOL REL GAP
Description:
Possible Values:
Default value:
1.0e-6
intpnt nl tol rel step
MSK DPAR INTPNT NL TOL REL STEP
Description:
Possible Values:
Any number between 1.0e-4 and 0.9999999.
Default value:
0.995
intpnt tol dfeas
MSK DPAR INTPNT TOL DFEAS
Description:
Possible Values:
Default value:
1.0e-8
intpnt tol dsafe
MSK DPAR INTPNT TOL DSAFE
Description:
Controls the initial dual starting point used by the interior-point optimizer. If the interior-
point optimizer converges slowly.
Possible Values:
Default value:
1.0
intpnt tol infeas
MSK DPAR INTPNT TOL INFEAS
Description:
Controls when the optimizer declares the model primal or dual infeasible. A small number
means the optimizer gets more conservative about declaring the model infeasible.
Possible Values:
Default value:
1.0e-8
intpnt tol mu red
MSK DPAR INTPNT TOL MU RED
Description:
Possible Values:
Default value:
1.0e-16
intpnt tol path
MSK DPAR INTPNT TOL PATH
Description:
Controls how close the interior-point optimizer follows the central path. A large value of
this parameter means the central is followed very closely. On numerical unstable problems
it may be worthwhile to increase this parameter.
Possible Values:
Default value:
1.0e-8
intpnt tol pfeas
MSK DPAR INTPNT TOL PFEAS
Description:
Possible Values:
Default value:
1.0e-8
intpnt tol psafe
MSK DPAR INTPNT TOL PSAFE
Description:
Controls the initial primal starting point used by the interior-point optimizer. If the interior-
point optimizer converges slowly and/or the constraint or variable bounds are very large,
then it may be worthwhile to increase this value.
Possible Values:
Default value:
1.0
intpnt tol rel gap
MSK DPAR INTPNT TOL REL GAP
Description:
Possible Values:
Default value:
1.0e-8
intpnt tol rel step
MSK DPAR INTPNT TOL REL STEP
Description:
Possible Values:
Default value:
0.9999
intpnt tol step size
MSK DPAR INTPNT TOL STEP SIZE
Description:
If the step size falls below the value of this parameter, then the interior-point optimizer
assumes that it is stalled. It it does not not make any progress.
Possible Values:
Default value:
1.0e-10
lower obj cut
MSK DPAR LOWER OBJ CUT
Description:
If either a primal or dual feasible solution is found proving that the optimal objective value
is outside, the interval [MSK DPAR LOWER OBJ CUT, MSK DPAR UPPER OBJ CUT], then MOSEK
is terminated.
Possible Values:
Default value:
-1.0e30
See also:
MSK DPAR LOWER OBJ CUT FINITE TRH Objective bound.
lower obj cut finite trh
MSK DPAR LOWER OBJ CUT FINITE TRH
Description:
If the lower objective cut is less than the value of this parameter value, then the lower
objective cut i.e. MSK DPAR LOWER OBJ CUT is treated as .
Possible Values:
Default value:
-0.5e30
mio disable term time
MSK DPAR MIO DISABLE TERM TIME
Description:
The termination criteria governed by
MSK IPAR MIO MAX NUM RELAXS
MSK IPAR MIO MAX NUM BRANCHES
is disabled the rst n seconds. This parameter species the number n. A negative value is
identical to innity i.e. the termination criteria are never checked.
Possible Values:
Default value:
-1.0
See also:
MSK IPAR MIO MAX NUM RELAXS Maximum number of relaxations in branch and bound search.
MSK IPAR MIO MAX NUM BRANCHES Maximum number of branches allowed during the branch
and bound search.
MSK DPAR MIO NEAR TOL ABS GAP Relaxed absolute optimality tolerance employed by the
mixed-integer optimizer.
MSK DPAR MIO NEAR TOL REL GAP The mixed-integer optimizer is terminated when this tol-
erance is satised.
mio heuristic time
MSK DPAR MIO HEURISTIC TIME
Description:
Minimum amount of time to be used in the heuristic search for a good feasible integer
solution. A negative values implies that the optimizer decides the amount of time to be
spent in the heuristic.
Possible Values:
Default value:
-1.0
mio max time
MSK DPAR MIO MAX TIME
Description:
This parameter limits the maximum time spent by the mixed-integer optimizer. A negative
number means innity.
Possible Values:
Default value:
-1.0
mio max time aprx opt
MSK DPAR MIO MAX TIME APRX OPT
Description:
Number of seconds spent by the mixed-integer optimizer before the MSK DPAR MIO TOL REL RELAX INT
is applied.
Possible Values:
Default value:
60
mio near tol abs gap
Description:
Relaxed absolute optimality tolerance employed by the mixed-integer optimizer. This ter-
mination criteria is delayed. See MSK DPAR MIO DISABLE TERM TIME for details.
Possible Values:
Default value:
0.0
See also:
MSK DPAR MIO DISABLE TERM TIME Certain termination criteria is disabled within the mixed-
integer optimizer for period time specied by the parameter.
mio near tol rel gap
Description:
The mixed-integer optimizer is terminated when this tolerance is satised. This termination
criteria is delayed. See MSK DPAR MIO DISABLE TERM TIME for details.
Possible Values:
Default value:
1.0e-3
See also:
mio rel add cut limited
MSK DPAR MIO REL ADD CUT LIMITED
Description:
Controls how many cuts the mixed-integer optimizer is allowed to add to the problem. Let
be the value of this parameter and m the number constraints, then mixed-integer optimizer
is allowed to m cuts.
Possible Values:
Default value:
0.75
mio rel gap const
MSK DPAR MIO REL GAP CONST
Description:
This value is used to compute the relative gap for the solution to an integer optimization
problem.
Possible Values:
Default value:
1.0e-10
mio tol abs gap
MSK DPAR MIO TOL ABS GAP
Description:
Possible Values:
Default value:
0.0
mio tol abs relax int
MSK DPAR MIO TOL ABS RELAX INT
Description:
Absolute relaxation tolerance of the integer constraints. I.e. min([x[ x|, x| [x[) is less
than the tolerance then the integer restrictions assumed to be satised.
Possible Values:
Default value:
1.0e-5
mio tol feas
MSK DPAR MIO TOL FEAS
Description:
Feasibility tolerance for mixed integer solver. Any solution with maximum infeasibility
below this value will be considered feasible.
Possible Values:
Default value:
1.0e-7
mio tol rel gap
MSK DPAR MIO TOL REL GAP
Description:
Possible Values:
Default value:
1.0e-4
mio tol rel relax int
MSK DPAR MIO TOL REL RELAX INT
Description:
Relative relaxation tolerance of the integer constraints. I.e (min([x[ x|, x| [x[)) is less
than the tolerance times [x[ then the integer restrictions assumed to be satised.
Possible Values:
Default value:
1.0e-6
mio tol x
MSK DPAR MIO TOL X
Description:
Possible Values:
Default value:
1.0e-6
nonconvex tol feas
MSK DPAR NONCONVEX TOL FEAS
Description:
Possible Values:
Default value:
1.0e-6
nonconvex tol opt
MSK DPAR NONCONVEX TOL OPT
Description:
Possible Values:
Default value:
1.0e-7
optimizer max time
MSK DPAR OPTIMIZER MAX TIME
Description:
Maximum amount of time the optimizer is allowed to spent on the optimization. A negative
number means innity.
Possible Values:
Default value:
-1.0
presolve tol aij
MSK DPAR PRESOLVE TOL AIJ
Description:
Absolute zero tolerance employed for a
ij
in the presolve.
Possible Values:
Default value:
1.0e-12
presolve tol lin dep
MSK DPAR PRESOLVE TOL LIN DEP
Description:
Possible Values:
Default value:
1.0e-6
presolve tol s
MSK DPAR PRESOLVE TOL S
Description:
Absolute zero tolerance employed for s
i
in the presolve.
Possible Values:
Default value:
1.0e-8
presolve tol x
MSK DPAR PRESOLVE TOL X
Description:
Absolute zero tolerance employed for x
j
in the presolve.
Possible Values:
Default value:
1.0e-8
qcqo reformulate rel drop tol
MSK DPAR QCQO REFORMULATE REL DROP TOL
Description:
This parameter determines when columns are dropped in incomplete cholesky factorization
doing reformulation of quadratic problems.
Possible Values:
Default value:
1e-15
sim lu tol rel piv
MSK DPAR SIM LU TOL REL PIV
Description:
Relative pivot tolerance employed when computing the LU factorization of the basis in the
simplex optimizers and in the basis identication procedure.
A value closer to 1.0 generally improves numerical stability but typically also implies an
increase in the computational work.
Possible Values:
Default value:
0.01
simplex abs tol piv
MSK DPAR SIMPLEX ABS TOL PIV
Description:
Possible Values:
Default value:
1.0e-7
upper obj cut
MSK DPAR UPPER OBJ CUT
Description:
If either a primal or dual feasible solution is found proving that the optimal objective value is
outside, [MSK DPAR LOWER OBJ CUT, MSK DPAR UPPER OBJ CUT], then MOSEK is terminated.
Possible Values:
Default value:
1.0e30
See also:
MSK DPAR UPPER OBJ CUT FINITE TRH Objective bound.
upper obj cut finite trh
MSK DPAR UPPER OBJ CUT FINITE TRH
Description:
If the upper objective cut is greater than the value of this value parameter, then the the
upper objective cut MSK DPAR UPPER OBJ CUT is treated as .
Possible Values:
Default value:
0.5e30
H.3 Integer parameters
MSK IPAR ALLOC ADD QNZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
MSK IPAR ANA SOL BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
H.3. INTEGER PARAMETERS 197
MSK IPAR ANA SOL PRINT VIOLATED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
MSK IPAR AUTO SORT A BEFORE OPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Controls whether the elements in each column of A are sorted before an optimization is performed.
MSK IPAR AUTO UPDATE SOL INFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Controls whether the solution information items are automatically updated after an optimization
is performed.
MSK IPAR BASIS SOLVE USE PLUS ONE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Controls the sign of the columns in the basis matrix corresponding to slack variables.
problem.
MSK IPAR CACHE LICENSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Control license caching.
MSK IPAR CONCURRENT NUM OPTIMIZERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
optimizer.
MSK IPAR CONCURRENT PRIORITY DUAL SIMPLEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
MSK IPAR CONCURRENT PRIORITY FREE SIMPLEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
MSK IPAR CONCURRENT PRIORITY INTPNT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
MSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
MSK IPAR CPU TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Species the CPU type.
MSK IPAR DATA CHECK. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Enable data checking for debug purposes.
MSK IPAR FEASREPAIR OPTIMIZE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
MSK IPAR INFEAS GENERIC NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
MSK IPAR INFEAS PREFER PRIMAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Controls which certicate is used if both primal- and dual- certicate of infeasibility is available.
MSK IPAR INTPNT DIFF STEP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
MSK IPAR INTPNT FACTOR DEBUG LVL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Controls factorization debug level.
MSK IPAR INTPNT FACTOR METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Controls the method used to factor the Newton equation system.
MSK IPAR INTPNT MAX NUM COR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Maximum number of correction steps.
MSK IPAR INTPNT MAX NUM REFINEMENT STEPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Maximum number of steps to be used by the iterative search direction renement.
MSK IPAR INTPNT NUM THREADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
number MOSEK will use this number of threads. If zero the number of threads used will equal
the number of cores detected on the machine.
MSK IPAR INTPNT OFF COL TRH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Controls the aggressiveness of the oending column detection.
MSK IPAR INTPNT ORDER METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Controls the ordering strategy.
MSK IPAR INTPNT REGULARIZATION USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
MSK IPAR INTPNT SCALING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
MSK IPAR INTPNT SOLVE FORM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
MSK IPAR INTPNT STARTING POINT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
MSK IPAR LIC TRH EXPIRY WRN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Controls when expiry warnings are issued.
MSK IPAR LICENSE ALLOW OVERUSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
MSK IPAR LICENSE DEBUG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Controls the license manager client debugging behavior.
MSK IPAR LICENSE PAUSE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
MSK IPAR LOG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
MSK IPAR LOG BI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
on.
shown.
MSK IPAR LOG FILE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
MSK IPAR LOG HEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
MSK IPAR LOG MIO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
MSK IPAR LOG SIM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
MSK IPAR LP WRITE IGNORE INCOMPATIBLE ITEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
MSK IPAR MAX NUM WARNINGS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
MSK IPAR MIO BRANCH DIR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
MSK IPAR MIO BRANCH PRIORITIES USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
MSK IPAR MIO CONSTRUCT SOL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Controls if an initial mixed integer solution should be constructed from the values of the integer
variables.
MSK IPAR MIO CUT LEVEL ROOT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Controls the cut level employed by the mixed-integer optimizer at the root node.
MSK IPAR MIO CUT LEVEL TREE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Controls the cut level employed by the mixed-integer optimizer in the tree.
MSK IPAR MIO FEASPUMP LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Controls the feasibility pump heuristic which is used to construct a good initial feasible solution.
MSK IPAR MIO HEURISTIC LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
Controls the heuristic employed by the mixed-integer optimizer to locate an initial integer feasible
solution.
MSK IPAR MIO HOTSTART . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
MSK IPAR MIO KEEP BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
Controls whether the integer presolve keeps bases in memory.
MSK IPAR MIO LOCAL BRANCH NUMBER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
MSK IPAR MIO MAX NUM RELAXS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Maximum number of relaxations in branch and bound search.
MSK IPAR MIO MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Turns on/o the mixed-integer mode.
MSK IPAR MIO NODE OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
MSK IPAR MIO NODE SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
MSK IPAR MIO OPTIMIZER MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
MSK IPAR MIO PRESOLVE AGGREGATE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Controls whether problem aggregation is performed in the mixed-integer presolve.
MSK IPAR MIO PRESOLVE PROBING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Controls whether probing is employed by the mixed-integer presolve.
MSK IPAR MIO PRESOLVE USE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
MSK IPAR MIO ROOT OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
MSK IPAR MIO STRONG BRANCH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
The depth from the root in which strong branching is employed.
MSK IPAR NONCONVEX MAX ITERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
MSK IPAR OBJECTIVE SENSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
If the objective sense for the task is undened, then the value of this parameter is used as the
default objective sense.
MSK IPAR OPF MAX TERMS PER LINE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
MSK IPAR OPF WRITE HEADER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
MSK IPAR OPF WRITE HINTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
MSK IPAR OPF WRITE PARAMETERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
MSK IPAR OPF WRITE PROBLEM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
MSK IPAR OPF WRITE SOL BAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
MSK IPAR OPF WRITE SOL ITG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
MSK IPAR OPF WRITE SOL ITR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
MSK IPAR OPF WRITE SOLUTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
MSK IPAR OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Controls which optimizer is used to optimize the task.
MSK IPAR PARAM READ CASE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
MSK IPAR PARAM READ IGN ERROR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
MSK IPAR PRESOLVE ELIM FILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Maximum amount of ll-in in the elimination phase.
MSK IPAR PRESOLVE ELIMINATOR MAX NUM TRIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
MSK IPAR PRESOLVE ELIMINATOR USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Currently not used.
MSK IPAR PRESOLVE LINDEP USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
MSK IPAR PRESOLVE LINDEP WORK LIM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Controls linear dependency check in presolve.
MSK IPAR QO SEPARABLE REFORMULATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
MSK IPAR READ ANZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
MSK IPAR READ CON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
MSK IPAR READ DATA COMPRESSED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Controls the input le decompression.
MSK IPAR READ DATA FORMAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
MSK IPAR READ KEEP FREE CON. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
MSK IPAR READ LP DROP NEW VARS IN BOU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Controls how the LP les are interpreted.
MSK IPAR READ LP QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
MSK IPAR READ MPS FORMAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
MSK IPAR READ MPS KEEP INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Controls if integer constraints are read.
MSK IPAR READ MPS OBJ SENSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
MSK IPAR READ MPS QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
MSK IPAR READ MPS RELAX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Controls the meaning of integer constraints.
MSK IPAR READ MPS WIDTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
MSK IPAR READ Q MODE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
MSK IPAR READ QNZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
MSK IPAR READ VAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
MSK IPAR SENSITIVITY ALL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Controls sensitivity report behavior.
MSK IPAR SENSITIVITY TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
MSK IPAR SIM BASIS FACTOR USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
ization sometimes improves the stability of the simplex optimizers, but in most cases there is a
performance penanlty.
MSK IPAR SIM DEGEN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
MSK IPAR SIM DUAL CRASH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
MSK IPAR SIM DUAL PHASEONE METHOD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
MSK IPAR SIM DUAL RESTRICT SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
MSK IPAR SIM DUAL SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Controls the dual simplex strategy.
MSK IPAR SIM EXPLOIT DUPVEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
MSK IPAR SIM HOTSTART . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
MSK IPAR SIM HOTSTART LU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Determines if the simplex optimizer should exploit the initial factorization.
MSK IPAR SIM INTEGER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
MSK IPAR SIM MAX NUM SETBACKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Controls how many set-backs that are allowed within a simplex optimizer.
MSK IPAR SIM NETWORK DETECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Level of aggressiveness of network detection.
MSK IPAR SIM NETWORK DETECT HOTSTART . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Level of aggressiveness of network detection in a simplex hot-start.
MSK IPAR SIM NETWORK DETECT METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
MSK IPAR SIM NON SINGULAR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
MSK IPAR SIM PRIMAL CRASH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Controls the simplex crash.
MSK IPAR SIM PRIMAL PHASEONE METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
MSK IPAR SIM PRIMAL RESTRICT SELECTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
MSK IPAR SIM PRIMAL SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Controls the primal simplex strategy.
MSK IPAR SIM REFACTOR FREQ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Controls the basis refactoring frequency.
MSK IPAR SIM REFORMULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
MSK IPAR SIM SAVE LU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Controls if the LU factorization stored should be replaced with the LU factorization corresponding
to the initial basis.
MSK IPAR SIM SCALING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
MSK IPAR SIM SCALING METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
MSK IPAR SIM SOLVE FORM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Controls whether the primal or the dual problem is solved by the primal-/dual- simplex optimizer.
MSK IPAR SIM STABILITY PRIORITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
MSK IPAR SIM SWITCH OPTIMIZER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Controls the simplex behavior.
MSK IPAR SOL FILTER KEEP BASIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
MSK IPAR SOL FILTER KEEP RANGED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Control the contents of the solution les.
MSK IPAR TIMING LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
MSK IPAR WARNING LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Warning level.
MSK IPAR WRITE DATA COMPRESSED. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Controls output le compression.
MSK IPAR WRITE DATA FORMAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Controls the output le format.
MSK IPAR WRITE DATA PARAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Controls output le data.
MSK IPAR WRITE FREE CON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
MSK IPAR WRITE GENERIC NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
MSK IPAR WRITE GENERIC NAMES IO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
MSK IPAR WRITE LP LINE WIDTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
MSK IPAR WRITE LP QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Controls LP output le format.
MSK IPAR WRITE LP STRICT FORMAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
MSK IPAR WRITE LP TERMS PER LINE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
MSK IPAR WRITE MPS INT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
MSK IPAR WRITE MPS OBJ SENSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
MSK IPAR WRITE MPS QUOTED NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
MSK IPAR WRITE MPS STRICT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Controls the output MPS le format.
MSK IPAR WRITE PRECISION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Controls data precision employed in when writing an MPS le.
MSK IPAR WRITE XML MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Controls if linear coecients should be written by row or column when writing in the XML le
format.
alloc add qnz
MSK IPAR ALLOC ADD QNZ
Description:
Additional number of Q non-zeros that are allocated space for when numanz exceeds maxnumqnz
during addition of new Q entries.
Possible Values:
Default value:
5000
ana sol basis
MSK IPAR ANA SOL BASIS
Description:
Possible values:
MSK ON Switch the option on.
MSK OFF Switch the option o.
Default value:
MSK ON
ana sol print violated
MSK IPAR ANA SOL PRINT VIOLATED
Description:
Possible values:
Default value:
MSK OFF
auto sort a before opt
MSK IPAR AUTO SORT A BEFORE OPT
Description:
Controls whether the elements in each column of A are sorted before an optimization is
performed. This is not required but makes the optimization more deterministic.
Possible values:
Default value:
MSK OFF
auto update sol info
MSK IPAR AUTO UPDATE SOL INFO
Description:
Controls whether the solution information items are automatically updated after an opti-
mization is performed.
Possible values:
Default value:
MSK ON
basis solve use plus one
MSK IPAR BASIS SOLVE USE PLUS ONE
Description:
If a slack variable is in the basis, then the corresponding column in the basis is a unit vector
with -1 in the right position. However, if this parameter is set to MSK ON, -1 is replaced by
1.
Possible values:
Default value:
MSK OFF
bi clean optimizer
MSK IPAR BI CLEAN OPTIMIZER
Description:
Possible values:
MSK OPTIMIZER INTPNT The interior-point optimizer is used.
MSK OPTIMIZER CONCURRENT The optimizer for nonconvex nonlinear problems.
MSK OPTIMIZER MIXED INT The mixed-integer optimizer.
MSK OPTIMIZER DUAL SIMPLEX The dual simplex optimizer is used.
MSK OPTIMIZER FREE The optimizer is chosen automatically.
MSK OPTIMIZER PRIMAL DUAL SIMPLEX The primal dual simplex optimizer is used.
MSK OPTIMIZER CONIC The optimizer for problems having conic constraints.
MSK OPTIMIZER NONCONVEX The optimizer for nonconvex nonlinear problems.
MSK OPTIMIZER QCONE For internal use only.
MSK OPTIMIZER PRIMAL SIMPLEX The primal simplex optimizer is used.
MSK OPTIMIZER FREE SIMPLEX One of the simplex optimizers is used.
Default value:
MSK OPTIMIZER FREE
bi ignore max iter
MSK IPAR BI IGNORE MAX ITER
Description:
If the parameter MSK IPAR INTPNT BASIS has the value MSK BI NO ERROR and the interior-
point optimizer has terminated due to maximum number of iterations, then basis identi-
cation is performed if this parameter has the value MSK ON.
Possible values:
Default value:
MSK OFF
bi ignore num error
MSK IPAR BI IGNORE NUM ERROR
Description:
If the parameter MSK IPAR INTPNT BASIS has the value MSK BI NO ERROR and the interior-
point optimizer has terminated due to a numerical problem, then basis identication is
performed if this parameter has the value MSK ON.
Possible values:
Default value:
MSK OFF
bi max iterations
MSK IPAR BI MAX ITERATIONS
Description:
Controls the maximum number of simplex iterations allowed to optimize a basis after the
basis identication.
Possible Values:
Default value:
1000000
cache license
MSK IPAR CACHE LICENSE
Description:
Species if the license is kept checked out for the lifetime of the mosek environment (on) or
returned to the server immediately after the optimization (o).
Check-in and check-out of licenses have an overhead. Frequent communication with the
license server should be avoided.
Possible values:
Default value:
MSK ON
cache size l1
MSK IPAR CACHE SIZE L1
Description:
Species the size of the cache of the computer. This parameter is potentially very important
for the eciency on computers if MOSEK cannot determine the cache size automatically.
If the cache size is negative, then MOSEK tries to determine the value automatically.
Possible Values:
Default value:
-1
cache size l2
MSK IPAR CACHE SIZE L2
Description:
Species the size of the cache of the computer. This parameter is potentially very important
for the eciency on computers where MOSEK cannot determine the cache size automati-
cally. If the cache size is negative, then MOSEK tries to determine the value automatically.
Possible Values:
Default value:
-1
check convexity
MSK IPAR CHECK CONVEXITY
Description:
Possible values:
MSK CHECK CONVEXITY SIMPLE Perform simple and fast convexity check.
MSK CHECK CONVEXITY NONE No convexity check.
MSK CHECK CONVEXITY FULL Perform a full convexity check.
Default value:
MSK CHECK CONVEXITY FULL
check task data
MSK IPAR CHECK TASK DATA
Description:
If this feature is turned on, then the task data is checked for bad values i.e. NaNs. before
an optimization is performed.
Possible values:
Default value:
MSK ON
concurrent num optimizers
MSK IPAR CONCURRENT NUM OPTIMIZERS
Description:
optimizer.
Possible Values:
Default value:
2
concurrent priority dual simplex
MSK IPAR CONCURRENT PRIORITY DUAL SIMPLEX
Description:
Possible Values:
Default value:
2
concurrent priority free simplex
MSK IPAR CONCURRENT PRIORITY FREE SIMPLEX
Description:
Possible Values:
Default value:
3
concurrent priority intpnt
MSK IPAR CONCURRENT PRIORITY INTPNT
Description:
Possible Values:
Default value:
4
concurrent priority primal simplex
MSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEX
Description:
Possible Values:
Default value:
1
cpu type
MSK IPAR CPU TYPE
Description:
Species the CPU type. By default MOSEK tries to auto detect the CPU type. Therefore,
we recommend to change this parameter only if the auto detection does not work properly.
Possible values:
MSK CPU POWERPC G5 A G5 PowerPC CPU.
MSK CPU INTEL PM An Intel PM cpu.
MSK CPU GENERIC An generic CPU type for the platform
MSK CPU UNKNOWN An unknown CPU.
MSK CPU AMD OPTERON An AMD Opteron (64 bit).
MSK CPU INTEL ITANIUM2 An Intel Itanium2.
MSK CPU AMD ATHLON An AMD Athlon.
MSK CPU HP PARISC20 An HP PA RISC version 2.0 CPU.
MSK CPU INTEL P4 An Intel Pentium P4 or Intel Xeon.
MSK CPU INTEL P3 An Intel Pentium P3.
MSK CPU INTEL CORE2 An Intel CORE2 cpu.
Default value:
MSK CPU UNKNOWN
data check
MSK IPAR DATA CHECK
Description:
If this option is turned on, then extensive data checking is enabled. It will slow down
MOSEK but on the other hand help locating bugs.
Possible values:
Default value:
MSK ON
feasrepair optimize
MSK IPAR FEASREPAIR OPTIMIZE
Description:
Possible values:
MSK FEASREPAIR OPTIMIZE NONE Do not optimize the feasibility repair problem.
MSK FEASREPAIR OPTIMIZE COMBINED Minimize with original objective subject to minimal
weighted violation of bounds.
MSK FEASREPAIR OPTIMIZE PENALTY Minimize weighted sum of violations.
Default value:
MSK FEASREPAIR OPTIMIZE NONE
infeas generic names
MSK IPAR INFEAS GENERIC NAMES
Description:
Controls whether generic names are used when an infeasible subproblem is created.
Possible values:
Default value:
MSK OFF
infeas prefer primal
MSK IPAR INFEAS PREFER PRIMAL
Description:
If both certicates of primal and dual infeasibility are supplied then only the primal is used
when this option is turned on.
Possible values:
Default value:
MSK ON
infeas report auto
MSK IPAR INFEAS REPORT AUTO
Description:
Controls whether an infeasibility report is automatically produced after the optimization if
the problem is primal or dual infeasible.
Possible values:
Default value:
MSK OFF
infeas report level
MSK IPAR INFEAS REPORT LEVEL
Description:
Controls the amount of information presented in an infeasibility report. Higher values imply
more information.
Possible Values:
Default value:
1
intpnt basis
MSK IPAR INTPNT BASIS
Description:
Controls whether the interior-point optimizer also computes an optimal basis.
Possible values:
MSK BI ALWAYS Basis identication is always performed even if the interior-point optimizer
terminates abnormally.
MSK BI NO ERROR Basis identication is performed if the interior-point optimizer terminates
without an error.
MSK BI NEVER Never do basis identication.
MSK BI IF FEASIBLE Basis identication is not performed if the interior-point optimizer
terminates with a problem status saying that the problem is primal or dual infeasible.
MSK BI OTHER Try another BI method.
Default value:
MSK BI ALWAYS
See also:
MSK IPAR BI IGNORE MAX ITER Turns on basis identication in case the interior-point opti-
mizer is terminated due to maximum number of iterations.
MSK IPAR BI IGNORE NUM ERROR Turns on basis identication in case the interior-point op-
timizer is terminated due to a numerical problem.
intpnt diff step
MSK IPAR INTPNT DIFF STEP
Description:
Possible values:
Default value:
MSK ON
intpnt factor debug lvl
MSK IPAR INTPNT FACTOR DEBUG LVL
Description:
Possible Values:
Default value:
0
intpnt factor method
MSK IPAR INTPNT FACTOR METHOD
Description:
Possible Values:
Default value:
0
intpnt max iterations
MSK IPAR INTPNT MAX ITERATIONS
Description:
Possible Values:
Default value:
400
intpnt max num cor
MSK IPAR INTPNT MAX NUM COR
Description:
Controls the maximum number of correctors allowed by the multiple corrector procedure.
A negative value means that MOSEK is making the choice.
Possible Values:
Any number between -1 and +inf.
Default value:
-1
intpnt max num refinement steps
MSK IPAR INTPNT MAX NUM REFINEMENT STEPS
Description:
Maximum number of steps to be used by the iterative renement of the search direction.
A negative value implies that the optimizer Chooses the maximum number of iterative
renement steps.
Possible Values:
Default value:
-1
intpnt num threads
MSK IPAR INTPNT NUM THREADS
Description:
number MOSEK will use this number of threads. If zero the number of threads used will
equal the number of cores detected on the machine.
Possible Values:
Any integer greater or equal to 0.
Default value:
1
intpnt off col trh
MSK IPAR INTPNT OFF COL TRH
Description:
Controls how many oending columns are detected in the Jacobian of the constraint matrix.
1 means aggressive detection, higher values mean less aggressive detection.
0 means no detection.
Possible Values:
Default value:
40
intpnt order method
MSK IPAR INTPNT ORDER METHOD
Description:
Controls the ordering strategy used by the interior-point optimizer when factorizing the
Newton equation system.
Possible values:
MSK ORDER METHOD NONE No ordering is used.
MSK ORDER METHOD APPMINLOC2 A variant of the approximate minimum local-ll-in ordering
is used.
MSK ORDER METHOD APPMINLOC1 Approximate minimum local-ll-in ordering is used.
MSK ORDER METHOD GRAPHPAR2 An alternative graph partitioning based ordering.
MSK ORDER METHOD FREE The ordering method is chosen automatically.
MSK ORDER METHOD GRAPHPAR1 Graph partitioning based ordering.
Default value:
MSK ORDER METHOD FREE
intpnt regularization use
MSK IPAR INTPNT REGULARIZATION USE
Description:
Possible values:
Default value:
MSK ON
intpnt scaling
MSK IPAR INTPNT SCALING
Description:
Possible values:
MSK SCALING NONE No scaling is performed.
MSK SCALING MODERATE A conservative scaling is performed.
MSK SCALING AGGRESSIVE A very aggressive scaling is performed.
MSK SCALING FREE The optimizer chooses the scaling heuristic.
Default value:
MSK SCALING FREE
intpnt solve form
MSK IPAR INTPNT SOLVE FORM
Description:
Possible values:
MSK SOLVE PRIMAL The optimizer should solve the primal problem.
MSK SOLVE DUAL The optimizer should solve the dual problem.
MSK SOLVE FREE The optimizer is free to solve either the primal or the dual problem.
Default value:
MSK SOLVE FREE
intpnt starting point
MSK IPAR INTPNT STARTING POINT
Description:
Possible values:
MSK STARTING POINT GUESS The optimizer guesses a starting point.
MSK STARTING POINT SATISFY BOUNDS The starting point is choosen to satisfy all the sim-
ple bounds on nonlinear variables. If this starting point is employed, then more care
than usual should employed when choosing the bounds on the nonlinear variables. In
particular very tight bounds should be avoided.
MSK STARTING POINT CONSTANT The optimizer constructs a starting point by assigning a
constant value to all primal and dual variables. This starting point is normally robust.
MSK STARTING POINT FREE The starting point is chosen automatically.
Default value:
MSK STARTING POINT FREE
lic trh expiry wrn
MSK IPAR LIC TRH EXPIRY WRN
Description:
If a license feature expires in a numbers days less than the value of this parameter then a
warning will be issued.
Possible Values:
Default value:
7
license allow overuse
MSK IPAR LICENSE ALLOW OVERUSE
Description:
Possible values:
Default value:
MSK ON
license cache time
MSK IPAR LICENSE CACHE TIME
Description:
Setting this parameter no longer has any eect. Please see MSK IPAR CACHE LICENSE for an
alternative.
Possible Values:
Any number between 0 and 65555.
Default value:
5
license check time
MSK IPAR LICENSE CHECK TIME
Description:
The parameter species the number of seconds between the checks of all the active licenses
in the MOSEK environment license cache. These checks are performed to determine if the
licenses should be returned to the server.
Possible Values:
Default value:
1
license debug
MSK IPAR LICENSE DEBUG
Description:
This option is used to turn on debugging of the incense manager.
Possible values:
Default value:
MSK OFF
license pause time
MSK IPAR LICENSE PAUSE TIME
Description:
If MSK IPAR LICENSE WAIT=MSK ON and no license is available, then MOSEK sleeps a num-
ber of milliseconds between each check of whether a license has become free.
Possible Values:
Default value:
100
license suppress expire wrns
MSK IPAR LICENSE SUPPRESS EXPIRE WRNS
Description:
Controls whether license features expire warnings are suppressed.
Possible values:
Default value:
MSK OFF
license wait
MSK IPAR LICENSE WAIT
Description:
If all licenses are in use MOSEK returns with an error code. However, by turning on this
parameter MOSEK will wait for an available license.
Possible values:
Default value:
MSK OFF
log
MSK IPAR LOG
Description:
Controls the amount of log information. The value 0 implies that all log information is
suppressed. A higher level implies that more information is logged.
Please note that if a task is employed to solve a sequence of optimization problems the value
of this parameter is reduced by the value of MSK IPAR LOG CUT SECOND OPT for the second
and any subsequent optimizations.
Possible Values:
Default value:
10
See also:
MSK IPAR LOG CUT SECOND OPT Controls the reduction in the log levels for the second and
any subsequent optimizations.
log bi
MSK IPAR LOG BI
Description:
Possible Values:
Default value:
4
log bi freq
MSK IPAR LOG BI FREQ
Description:
Controls how frequent the optimizer outputs information about the basis identication and
how frequent the user-dened call-back function is called.
Possible Values:
Default value:
2500
log check convexity
MSK IPAR LOG CHECK CONVEXITY
Description:
Controls logging in convexity check on quadratic problems. Set to a positive value to turn
logging on.
If a quadratic coecient matrix is found to violate the requirement of PSD (NSD) then a
list of negative (positive) pivot elements is printed. The absolute value of the pivot elements
is also shown.
Possible Values:
Default value:
0
log concurrent
MSK IPAR LOG CONCURRENT
Description:
Possible Values:
Default value:
1
log cut second opt
MSK IPAR LOG CUT SECOND OPT
Description:
If a task is employed to solve a sequence of optimization problems, then the value of the log
levels is reduced by the value of this parameter. E.g MSK IPAR LOG and MSK IPAR LOG SIM
are reduced by the value of this parameter for the second and any subsequent optimizations.
Possible Values:
Default value:
1
See also:
MSK IPAR LOG Controls the amount of log information.
MSK IPAR LOG INTPNT Controls the amount of log information from the interior-point opti-
mizers.
MSK IPAR LOG MIO Controls the amount of log information from the mixed-integer optimiz-
ers.
MSK IPAR LOG SIM Controls the amount of log information from the simplex optimizers.
log factor
MSK IPAR LOG FACTOR
Description:
Possible Values:
Default value:
1
log feasrepair
MSK IPAR LOG FEASREPAIR
Description:
Possible Values:
Default value:
0
log file
MSK IPAR LOG FILE
Description:
Possible Values:
Default value:
1
log head
MSK IPAR LOG HEAD
Description:
Possible Values:
Default value:
1
log infeas ana
MSK IPAR LOG INFEAS ANA
Description:
Controls amount of output printed by the infeasibility analyzer procedures. A higher level
Possible Values:
Default value:
1
log intpnt
MSK IPAR LOG INTPNT
Description:
Controls amount of output printed printed by the interior-point optimizer. A higher level
Possible Values:
Default value:
4
log mio
MSK IPAR LOG MIO
Description:
Controls the log level for the mixed-integer optimizer. A higher level implies that more
Possible Values:
Default value:
4
log mio freq
MSK IPAR LOG MIO FREQ
Description:
Controls how frequent the mixed-integer optimizer prints the log line. It will print line every
time MSK IPAR LOG MIO FREQ relaxations have been solved.
Possible Values:
A integer value.
Default value:
1000
log nonconvex
MSK IPAR LOG NONCONVEX
Description:
Possible Values:
Default value:
1
log optimizer
MSK IPAR LOG OPTIMIZER
Description:
Possible Values:
Default value:
1
log order
MSK IPAR LOG ORDER
Description:
Possible Values:
Default value:
1
log param
MSK IPAR LOG PARAM
Description:
Possible Values:
Default value:
0
log presolve
MSK IPAR LOG PRESOLVE
Description:
Controls amount of output printed by the presolve procedure. A higher level implies that
Possible Values:
Default value:
1
log response
MSK IPAR LOG RESPONSE
Description:
Controls amount of output printed when response codes are reported. A higher level implies
that more information is logged.
Possible Values:
Default value:
0
log sensitivity
MSK IPAR LOG SENSITIVITY
Description:
Controls the amount of logging during the sensitivity analysis. 0: Means no logging infor-
mation is produced. 1: Timing information is printed. 2: Sensitivity results are printed.
Possible Values:
Default value:
1
log sensitivity opt
MSK IPAR LOG SENSITIVITY OPT
Description:
Controls the amount of logging from the optimizers employed during the sensitivity analysis.
0 means no logging information is produced.
Possible Values:
Default value:
0
log sim
MSK IPAR LOG SIM
Description:
Controls amount of output printed by the simplex optimizer. A higher level implies that
Possible Values:
Default value:
4
log sim freq
MSK IPAR LOG SIM FREQ
Description:
Controls how frequent the simplex optimizer outputs information about the optimization
and how frequent the user-dened call-back function is called.
Possible Values:
Default value:
500
log sim minor
MSK IPAR LOG SIM MINOR
Description:
Possible Values:
Default value:
1
log sim network freq
MSK IPAR LOG SIM NETWORK FREQ
Description:
Controls how frequent the network simplex optimizer outputs information about the opti-
mization and how frequent the user-dened call-back function is called. The network opti-
mizer will use a logging frequency equal to MSK IPAR LOG SIM FREQ times MSK IPAR LOG SIM NETWORK FREQ.
Possible Values:
Default value:
50
log storage
MSK IPAR LOG STORAGE
Description:
When turned on, MOSEK prints messages regarding the storage usage and allocation.
Possible Values:
Default value:
0
lp write ignore incompatible items
MSK IPAR LP WRITE IGNORE INCOMPATIBLE ITEMS
Description:
Possible values:
Default value:
MSK OFF
max num warnings
MSK IPAR MAX NUM WARNINGS
Description:
Possible Values:
Default value:
10
mio branch dir
MSK IPAR MIO BRANCH DIR
Description:
Possible values:
MSK BRANCH DIR DOWN The mixed-integer optimizer always chooses the down branch rst.
MSK BRANCH DIR UP The mixed-integer optimizer always chooses the up branch rst.
MSK BRANCH DIR FREE The mixed-integer optimizer decides which branch to choose.
Default value:
MSK BRANCH DIR FREE
mio branch priorities use
MSK IPAR MIO BRANCH PRIORITIES USE
Description:
Possible values:
Default value:
MSK ON
mio construct sol
MSK IPAR MIO CONSTRUCT SOL
Description:
If set to MSK ON and all integer variables have been given a value for which a feasible mixed
integer solution exists, then MOSEK generates an initial solution to the mixed integer
problem by xing all integer values and solving the remaining problem.
Possible values:
Default value:
MSK OFF
mio cont sol
MSK IPAR MIO CONT SOL
Description:
Controls the meaning of the interior-point and basic solutions in mixed integer problems.
Possible values:
MSK MIO CONT SOL ITG The reported interior-point and basic solutions are a solution to the
problem with all integer variables xed at the value they have in the integer solution.
A solution is only reported in case the problem has a primal feasible solution.
MSK MIO CONT SOL NONE No interior-point or basic solution are reported when the mixed-
integer optimizer is used.
MSK MIO CONT SOL ROOT The reported interior-point and basic solutions are a solution to
the root node problem when mixed-integer optimizer is used.
MSK MIO CONT SOL ITG REL In case the problem is primal feasible then the reported interior-
point and basic solutions are a solution to the problem with all integer variables xed
at the value they have in the integer solution. If the problem is primal infeasible, then
the solution to the root node problem is reported.
Default value:
MSK MIO CONT SOL NONE
mio cut level root
MSK IPAR MIO CUT LEVEL ROOT
Description:
Controls the cut level employed by the mixed-integer optimizer at the root node. A negative
value means a default value determined by the mixed-integer optimizer is used. By adding
the appropriate values from the following table the employed cut types can be controlled.
GUB cover +2
Flow cover +4
Lifting +8
Plant location +16
Disaggregation +32
Knapsack cover +64
Lattice +128
Gomory +256
Coecient reduction +512
GCD +1024
Obj. integrality +2048
Possible Values:
Any value.
Default value:
-1
mio cut level tree
MSK IPAR MIO CUT LEVEL TREE
Description:
Controls the cut level employed by the mixed-integer optimizer at the tree. See MSK IPAR MIO CUT LEVEL ROOT
for an explanation of the parameter values.
Possible Values:
Any value.
Default value:
-1
mio feaspump level
MSK IPAR MIO FEASPUMP LEVEL
Description:
Feasibility pump is a heuristic designed to compute an initial feasible solution. A value of
0 implies that the feasibility pump heuristic is not used. A value of -1 implies that the
mixed-integer optimizer decides how the feasibility pump heuristic is used. A larger value
than 1 implies that the feasibility pump is employed more aggressively. Normally a value
beyond 3 is not worthwhile.
Possible Values:
Any number between -inf and 3.
Default value:
-1
mio heuristic level
MSK IPAR MIO HEURISTIC LEVEL
Description:
Controls the heuristic employed by the mixed-integer optimizer to locate an initial good
integer feasible solution. A value of zero means the heuristic is not used at all. A larger value
than 0 means that a gradually more sophisticated heuristic is used which is computationally
more expensive. A negative value implies that the optimizer chooses the heuristic. Normally
a value around 3 to 5 should be optimal.
Possible Values:
Any value.
Default value:
-1
mio hotstart
MSK IPAR MIO HOTSTART
Description:
Possible values:
Default value:
MSK ON
mio keep basis
MSK IPAR MIO KEEP BASIS
Description:
Controls whether the integer presolve keeps bases in memory. This speeds on the solution
process at cost of bigger memory consumption.
Possible values:
Default value:
MSK ON
mio local branch number
MSK IPAR MIO LOCAL BRANCH NUMBER
Description:
Possible Values:
Default value:
-1
mio max num branches
Description:
Maximum number of branches allowed during the branch and bound search. A negative
value means innite.
Possible Values:
Default value:
-1
See also:
mio max num relaxs
Description:
Maximum number of relaxations allowed during the branch and bound search. A negative
value means innite.
Possible Values:
Default value:
-1
See also:
mio max num solutions
MSK IPAR MIO MAX NUM SOLUTIONS
Description:
The mixed-integer optimizer can be terminated after a certain number of dierent feasible
solutions has been located. If this parameter has the value n and n is strictly positive, then
the mixed-integer optimizer will be terminated when n feasible solutions have been located.
Possible Values:
Default value:
-1
See also:
mio mode
MSK IPAR MIO MODE
Description:
Controls whether the optimizer includes the integer restrictions when solving a (mixed)
integer optimization problem.
Possible values:
MSK MIO MODE IGNORED The integer constraints are ignored and the problem is solved as a
continuous problem.
MSK MIO MODE LAZY Integer restrictions should be satised if an optimizer is available for
the problem.
MSK MIO MODE SATISFIED Integer restrictions should be satised.
Default value:
MSK MIO MODE SATISFIED
mio node optimizer
MSK IPAR MIO NODE OPTIMIZER
Description:
Possible values:
Default value:
MSK OPTIMIZER FREE
mio node selection
MSK IPAR MIO NODE SELECTION
Description:
Possible values:
MSK MIO NODE SELECTION PSEUDO The optimizer employs selects the node based on a pseudo
cost estimate.
MSK MIO NODE SELECTION HYBRID The optimizer employs a hybrid strategy.
MSK MIO NODE SELECTION FREE The optimizer decides the node selection strategy.
MSK MIO NODE SELECTION WORST The optimizer employs a worst bound node selection strat-
egy.
MSK MIO NODE SELECTION BEST The optimizer employs a best bound node selection strat-
egy.
MSK MIO NODE SELECTION FIRST The optimizer employs a depth rst node selection strat-
egy.
Default value:
MSK MIO NODE SELECTION FREE
mio optimizer mode
MSK IPAR MIO OPTIMIZER MODE
Description:
Possible Values:
Default value:
0
mio presolve aggregate
MSK IPAR MIO PRESOLVE AGGREGATE
Description:
Controls whether the presolve used by the mixed-integer optimizer tries to aggregate the
constraints.
Possible values:
Default value:
MSK ON
mio presolve probing
MSK IPAR MIO PRESOLVE PROBING
Description:
Controls whether the mixed-integer presolve performs probing. Probing can be very time
consuming.
Possible values:
Default value:
MSK ON
mio presolve use
MSK IPAR MIO PRESOLVE USE
Description:
Possible values:
Default value:
MSK ON
mio root optimizer
MSK IPAR MIO ROOT OPTIMIZER
Description:
Possible values:
Default value:
MSK OPTIMIZER FREE
mio strong branch
MSK IPAR MIO STRONG BRANCH
Description:
The value species the depth from the root in which strong branching is used. A negative
value means that the optimizer chooses a default value automatically.
Possible Values:
Default value:
-1
nonconvex max iterations
MSK IPAR NONCONVEX MAX ITERATIONS
Description:
Possible Values:
Default value:
100000
objective sense
MSK IPAR OBJECTIVE SENSE
Description:
If the objective sense for the task is undened, then the value of this parameter is used as
the default objective sense.
Possible values:
MSK OBJECTIVE SENSE MINIMIZE The problem should be minimized.
MSK OBJECTIVE SENSE UNDEFINED The objective sense is undened.
MSK OBJECTIVE SENSE MAXIMIZE The problem should be maximized.
Default value:
MSK OBJECTIVE SENSE MINIMIZE
opf max terms per line
MSK IPAR OPF MAX TERMS PER LINE
Description:
Possible Values:
Default value:
5
opf write header
MSK IPAR OPF WRITE HEADER
Description:
Possible values:
Default value:
MSK ON
opf write hints
MSK IPAR OPF WRITE HINTS
Description:
Possible values:
Default value:
MSK ON
opf write parameters
MSK IPAR OPF WRITE PARAMETERS
Description:
Possible values:
Default value:
MSK OFF
opf write problem
MSK IPAR OPF WRITE PROBLEM
Description:
Possible values:
Default value:
MSK ON
opf write sol bas
MSK IPAR OPF WRITE SOL BAS
Description:
If MSK IPAR OPF WRITE SOLUTIONS is MSK ON and a basic solution is dened, include the
basic solution in OPF les.
Possible values:
Default value:
MSK ON
opf write sol itg
MSK IPAR OPF WRITE SOL ITG
Description:
If MSK IPAR OPF WRITE SOLUTIONS is MSK ON and an integer solution is dened, write the
integer solution in OPF les.
Possible values:
Default value:
MSK ON
opf write sol itr
MSK IPAR OPF WRITE SOL ITR
Description:
If MSK IPAR OPF WRITE SOLUTIONS is MSK ON and an interior solution is dened, write the
interior solution in OPF les.
Possible values:
Default value:
MSK ON
opf write solutions
MSK IPAR OPF WRITE SOLUTIONS
Description:
Possible values:
Default value:
MSK OFF
optimizer
MSK IPAR OPTIMIZER
Description:
The paramter controls which optimizer is used to optimize the task.
Possible values:
Default value:
MSK OPTIMIZER FREE
param read case name
MSK IPAR PARAM READ CASE NAME
Description:
Possible values:
Default value:
MSK ON
param read ign error
MSK IPAR PARAM READ IGN ERROR
Description:
Possible values:
Default value:
MSK OFF
presolve elim fill
MSK IPAR PRESOLVE ELIM FILL
Description:
Controls the maximum amount of ll-in that can be created during the elimination phase
of the presolve. This parameter times (numcon+numvar) denotes the amount of ll-in.
Possible Values:
Default value:
1
presolve eliminator max num tries
MSK IPAR PRESOLVE ELIMINATOR MAX NUM TRIES
Description:
Possible Values:
A negative value implies MOSEK decides maximum number of times.
Default value:
-1
presolve eliminator use
MSK IPAR PRESOLVE ELIMINATOR USE
Description:
Possible values:
Default value:
MSK ON
presolve level
MSK IPAR PRESOLVE LEVEL
Description:
Currently not used.
Possible Values:
Default value:
-1
presolve lindep use
MSK IPAR PRESOLVE LINDEP USE
Description:
Possible values:
Default value:
MSK ON
presolve lindep work lim
MSK IPAR PRESOLVE LINDEP WORK LIM
Description:
Is used to limit the amount of work that can done to locate linear dependencies. In general
the higher value this parameter is given the less work can be used. However, a value of 0
means no limit on the amount work that can be used.
Possible Values:
Default value:
1
presolve use
MSK IPAR PRESOLVE USE
Description:
Possible values:
MSK PRESOLVE MODE ON The problem is presolved before it is optimized.
MSK PRESOLVE MODE OFF The problem is not presolved before it is optimized.
MSK PRESOLVE MODE FREE It is decided automatically whether to presolve before the prob-
lem is optimized.
Default value:
MSK PRESOLVE MODE FREE
qo separable reformulation
MSK IPAR QO SEPARABLE REFORMULATION
Description:
Possible values:
Default value:
MSK OFF
read add anz
MSK IPAR READ ADD ANZ
Description:
Additional number of non-zeros in A that is made room for in the problem.
Possible Values:
Default value:
0
read add con
MSK IPAR READ ADD CON
Description:
Possible Values:
Default value:
0
read add cone
MSK IPAR READ ADD CONE
Description:
Possible Values:
Default value:
0
read add qnz
MSK IPAR READ ADD QNZ
Description:
Additional number of non-zeros in the Q matrices that is made room for in the problem.
Possible Values:
Default value:
0
read add var
MSK IPAR READ ADD VAR
Description:
Possible Values:
Default value:
0
read anz
MSK IPAR READ ANZ
Description:
Expected maximum number of A non-zeros to be read. The option is used only by fast
MPS and LP le readers.
Possible Values:
Default value:
100000
read con
MSK IPAR READ CON
Description:
Expected maximum number of constraints to be read. The option is only used by fast MPS
and LP le readers.
Possible Values:
Default value:
10000
read cone
MSK IPAR READ CONE
Description:
Expected maximum number of conic constraints to be read. The option is used only by fast
MPS and LP le readers.
Possible Values:
Default value:
2500
read data compressed
MSK IPAR READ DATA COMPRESSED
Description:
If this option is turned on,it is assumed that the data le is compressed.
Possible values:
Default value:
MSK OFF
read data format
MSK IPAR READ DATA FORMAT
Description:
Possible values:
MSK DATA FORMAT XML The data le is an XML formatted le.
MSK DATA FORMAT FREE MPS The data data a free MPS formatted le.
MSK DATA FORMAT EXTENSION The le extension is used to determine the data le format.
MSK DATA FORMAT MPS The data le is MPS formatted.
MSK DATA FORMAT LP The data le is LP formatted.
MSK DATA FORMAT MBT The data le is a MOSEK binary task le.
MSK DATA FORMAT OP The data le is an optimization problem formatted le.
Default value:
MSK DATA FORMAT EXTENSION
read keep free con
MSK IPAR READ KEEP FREE CON
Description:
Possible values:
Default value:
MSK OFF
read lp drop new vars in bou
MSK IPAR READ LP DROP NEW VARS IN BOU
Description:
If this option is turned on, MOSEK will drop variables that are dened for the rst time in
the bounds section.
Possible values:
Default value:
MSK OFF
read lp quoted names
MSK IPAR READ LP QUOTED NAMES
Description:
Possible values:
Default value:
MSK ON
read mps format
MSK IPAR READ MPS FORMAT
Description:
Possible values:
MSK MPS FORMAT STRICT It is assumed that the input le satises the MPS format strictly.
MSK MPS FORMAT RELAXED It is assumed that the input le satises a slightly relaxed version
of the MPS format.
MSK MPS FORMAT FREE It is assumed that the input le satises the free MPS format. This
implies that spaces are not allowed in names. Otherwise the format is free.
Default value:
MSK MPS FORMAT RELAXED
read mps keep int
MSK IPAR READ MPS KEEP INT
Description:
Controls whether MOSEK should keep the integer restrictions on the variables while reading
the MPS le.
Possible values:
Default value:
MSK ON
read mps obj sense
MSK IPAR READ MPS OBJ SENSE
Description:
If turned on, the MPS reader uses the objective sense section. Otherwise the MPS reader
ignores it.
Possible values:
Default value:
MSK ON
read mps quoted names
MSK IPAR READ MPS QUOTED NAMES
Description:
If a name is in quotes when reading an MPS le, then the quotes will be removed.
Possible values:
Default value:
MSK ON
read mps relax
MSK IPAR READ MPS RELAX
Description:
If this option is turned on, then mixed integer constraints are ignored when a problem is
read.
Possible values:
Default value:
MSK ON
read mps width
MSK IPAR READ MPS WIDTH
Description:
Possible Values:
Any positive number greater than 80.
Default value:
1024
read q mode
MSK IPAR READ Q MODE
Description:
Possible values:
MSK Q READ ADD All elements in a Q matrix are assumed to belong to the lower triangular
part. Duplicate elements in a Q matrix are added together.
MSK Q READ DROP LOWER All elements in the strict lower triangular part of the Q matrices
are dropped.
MSK Q READ DROP UPPER All elements in the strict upper triangular part of the Q matrices
are dropped.
Default value:
MSK Q READ ADD
read qnz
MSK IPAR READ QNZ
Description:
Expected maximum number of Q non-zeros to be read. The option is used only by MPS
and LP le readers.
Possible Values:
Default value:
20000
read task ignore param
MSK IPAR READ TASK IGNORE PARAM
Description:
Controls whether MOSEK should ignore the parameter setting dened in the task le and
use the default parameter setting instead.
Possible values:
Default value:
MSK OFF
read var
MSK IPAR READ VAR
Description:
Expected maximum number of variable to be read. The option is used only by MPS and
LP le readers.
Possible Values:
Default value:
10000
sensitivity all
MSK IPAR SENSITIVITY ALL
Description:
Not applicable.
Possible values:
Default value:
MSK OFF
sensitivity optimizer
MSK IPAR SENSITIVITY OPTIMIZER
Description:
Possible values:
Default value:
MSK OPTIMIZER FREE SIMPLEX
sensitivity type
MSK IPAR SENSITIVITY TYPE
Description:
Possible values:
MSK SENSITIVITY TYPE OPTIMAL PARTITION Optimal partition sensitivity analysis is per-
formed.
MSK SENSITIVITY TYPE BASIS Basis sensitivity analysis is performed.
Default value:
MSK SENSITIVITY TYPE BASIS
sim basis factor use
MSK IPAR SIM BASIS FACTOR USE
Description:
ization sometimes improves the stability of the simplex optimizers, but in most cases there
is a performance penanlty.
Possible values:
Default value:
MSK ON
sim degen
MSK IPAR SIM DEGEN
Description:
Possible values:
MSK SIM DEGEN NONE The simplex optimizer should use no degeneration strategy.
MSK SIM DEGEN MODERATE The simplex optimizer should use a moderate degeneration strat-
egy.
MSK SIM DEGEN MINIMUM The simplex optimizer should use a minimum degeneration strat-
egy.
MSK SIM DEGEN AGGRESSIVE The simplex optimizer should use an aggressive degeneration
strategy.
MSK SIM DEGEN FREE The simplex optimizer chooses the degeneration strategy.
Default value:
MSK SIM DEGEN FREE
sim dual crash
MSK IPAR SIM DUAL CRASH
Description:
In general if a basis consists of more than (100-this parameter value)% xed variables, then
a crash will be performed.
Possible Values:
Default value:
90
sim dual phaseone method
MSK IPAR SIM DUAL PHASEONE METHOD
Description:
Possible Values:
Default value:
0
sim dual restrict selection
MSK IPAR SIM DUAL RESTRICT SELECTION
Description:
The dual simplex optimizer can use a so-called restricted selection/pricing strategy to
chooses the outgoing variable. Hence, if restricted selection is applied, then the dual simplex
optimizer rst choose a subset of all the potential outgoing variables. Next, for some time
it will choose the outgoing variable only among the subset. From time to time the subset
is redened.
A larger value of this parameter implies that the optimizer will be more aggressive in its
restriction strategy, i.e. a value of 0 implies that the restriction strategy is not applied at
all.
Possible Values:
Default value:
50
sim dual selection
MSK IPAR SIM DUAL SELECTION
Description:
Controls the choice of the incoming variable, known as the selection strategy, in the dual
simplex optimizer.
Possible values:
MSK SIM SELECTION FULL The optimizer uses full pricing.
MSK SIM SELECTION PARTIAL The optimizer uses a partial selection approach. The ap-
proach is usually benecial if the number of variables is much larger than the number
of constraints.
MSK SIM SELECTION FREE The optimizer chooses the pricing strategy.
MSK SIM SELECTION ASE The optimizer uses approximate steepest-edge pricing.
MSK SIM SELECTION DEVEX The optimizer uses devex steepest-edge pricing (or if it is not
available an approximate steep-edge selection).
MSK SIM SELECTION SE The optimizer uses steepest-edge selection (or if it is not available
an approximate steep-edge selection).
Default value:
MSK SIM SELECTION FREE
sim exploit dupvec
MSK IPAR SIM EXPLOIT DUPVEC
Description:
Possible values:
MSK SIM EXPLOIT DUPVEC ON Allow the simplex optimizer to exploit duplicated columns.
MSK SIM EXPLOIT DUPVEC OFF Disallow the simplex optimizer to exploit duplicated columns.
MSK SIM EXPLOIT DUPVEC FREE The simplex optimizer can choose freely.
Default value:
MSK SIM EXPLOIT DUPVEC OFF
sim hotstart
MSK IPAR SIM HOTSTART
Description:
Possible values:
MSK SIM HOTSTART NONE The simplex optimizer performs a coldstart.
MSK SIM HOTSTART STATUS KEYS Only the status keys of the constraints and variables are
used to choose the type of hot-start.
MSK SIM HOTSTART FREE The simplex optimize chooses the hot-start type.
Default value:
MSK SIM HOTSTART FREE
sim hotstart lu
MSK IPAR SIM HOTSTART LU
Description:
Determines if the simplex optimizer should exploit the initial factorization.
Possible values:
Default value:
MSK ON
sim integer
MSK IPAR SIM INTEGER
Description:
Possible Values:
Default value:
0
sim max iterations
MSK IPAR SIM MAX ITERATIONS
Description:
Possible Values:
Default value:
10000000
sim max num setbacks
MSK IPAR SIM MAX NUM SETBACKS
Description:
Controls how many set-backs are allowed within a simplex optimizer. A set-back is an event
where the optimizer moves in the wrong direction. This is impossible in theory but may
happen due to numerical problems.
Possible Values:
Default value:
250
sim network detect
MSK IPAR SIM NETWORK DETECT
Description:
The simplex optimizer is capable of exploiting a network ow component in a problem.
However it is only worthwhile to exploit the network ow component if it is suciently
large. This parameter controls how large the network component has to be in relative
terms before it is exploited. For instance a value of 20 means at least 20% of the model
should be a network before it is exploited. If this value is larger than 100 the network ow
component is never detected or exploited.
Possible Values:
Default value:
101
sim network detect hotstart
MSK IPAR SIM NETWORK DETECT HOTSTART
Description:
This parameter controls has large the network component in relative terms has to be
before it is exploited in a simplex hot-start. The network component should be equal or
larger than
max(MSK_IPAR_SIM_NETWORK_DETECT,MSK_IPAR_SIM_NETWORK_DETECT_HOTSTART)
before it is exploited. If this value is larger than 100 the network ow component is never
detected or exploited.
Possible Values:
Default value:
100
sim network detect method
MSK IPAR SIM NETWORK DETECT METHOD
Description:
Possible values:
MSK NETWORK DETECT SIMPLE The network detection should use a very simple heuristic.
MSK NETWORK DETECT ADVANCED The network detection should use a more advanced heuris-
tic.
MSK NETWORK DETECT FREE The network detection is free.
Default value:
MSK NETWORK DETECT FREE
sim non singular
MSK IPAR SIM NON SINGULAR
Description:
Possible values:
Default value:
MSK ON
sim primal crash
MSK IPAR SIM PRIMAL CRASH
Description:
Controls whether crashing is performed in the primal simplex optimizer.
In general, if a basis consists of more than (100-this parameter value)% xed variables, then
a crash will be performed.
Possible Values:
Any nonnegative integer value.
Default value:
90
sim primal phaseone method
MSK IPAR SIM PRIMAL PHASEONE METHOD
Description:
Possible Values:
Default value:
0
sim primal restrict selection
MSK IPAR SIM PRIMAL RESTRICT SELECTION
Description:
The primal simplex optimizer can use a so-called restricted selection/pricing strategy to
chooses the outgoing variable. Hence, if restricted selection is applied, then the primal
simplex optimizer rst choose a subset of all the potential incoming variables. Next, for
some time it will choose the incoming variable only among the subset. From time to time
the subset is redened.
A larger value of this parameter implies that the optimizer will be more aggressive in its
restriction strategy, i.e. a value of 0 implies that the restriction strategy is not applied at
all.
Possible Values:
Default value:
50
sim primal selection
MSK IPAR SIM PRIMAL SELECTION
Description:
Controls the choice of the incoming variable, known as the selection strategy, in the primal
simplex optimizer.
Possible values:
MSK SIM SELECTION FULL The optimizer uses full pricing.
MSK SIM SELECTION PARTIAL The optimizer uses a partial selection approach. The ap-
proach is usually benecial if the number of variables is much larger than the number
of constraints.
MSK SIM SELECTION FREE The optimizer chooses the pricing strategy.
MSK SIM SELECTION ASE The optimizer uses approximate steepest-edge pricing.
MSK SIM SELECTION DEVEX The optimizer uses devex steepest-edge pricing (or if it is not
available an approximate steep-edge selection).
MSK SIM SELECTION SE The optimizer uses steepest-edge selection (or if it is not available
Default value:
MSK SIM SELECTION FREE
sim refactor freq
MSK IPAR SIM REFACTOR FREQ
Description:
Controls how frequent the basis is refactorized. The value 0 means that the optimizer
determines the best point of refactorization.
It is strongly recommended NOT to change this parameter.
Possible Values:
Default value:
0
sim reformulation
MSK IPAR SIM REFORMULATION
Description:
Possible values:
MSK SIM REFORMULATION ON Allow the simplex optimizer to reformulate the problem.
MSK SIM REFORMULATION AGGRESSIVE The simplex optimizer should use an aggressive re-
formulation strategy.
MSK SIM REFORMULATION OFF Disallow the simplex optimizer to reformulate the problem.
MSK SIM REFORMULATION FREE The simplex optimizer can choose freely.
Default value:
MSK SIM REFORMULATION OFF
sim save lu
MSK IPAR SIM SAVE LU
Description:
Controls if the LU factorization stored should be replaced with the LU factorization corre-
sponding to the initial basis.
Possible values:
Default value:
MSK OFF
sim scaling
MSK IPAR SIM SCALING
Description:
Possible values:
MSK SCALING NONE No scaling is performed.
MSK SCALING MODERATE A conservative scaling is performed.
MSK SCALING AGGRESSIVE A very aggressive scaling is performed.
MSK SCALING FREE The optimizer chooses the scaling heuristic.
Default value:
MSK SCALING FREE
sim scaling method
MSK IPAR SIM SCALING METHOD
Description:
Possible values:
MSK SCALING METHOD POW2 Scales only with power of 2 leaving the mantissa untouched.
MSK SCALING METHOD FREE The optimizer chooses the scaling heuristic.
Default value:
MSK SCALING METHOD POW2
sim solve form
MSK IPAR SIM SOLVE FORM
Description:
Controls whether the primal or the dual problem is solved by the primal-/dual- simplex
optimizer.
Possible values:
MSK SOLVE PRIMAL The optimizer should solve the primal problem.
MSK SOLVE DUAL The optimizer should solve the dual problem.
MSK SOLVE FREE The optimizer is free to solve either the primal or the dual problem.
Default value:
MSK SOLVE FREE
sim stability priority
MSK IPAR SIM STABILITY PRIORITY
Description:
Possible Values:
Default value:
50
sim switch optimizer
MSK IPAR SIM SWITCH OPTIMIZER
Description:
The simplex optimizer sometimes chooses to solve the dual problem instead of the primal
problem. This implies that if you have chosen to use the dual simplex optimizer and the
problem is dualized, then it actually makes sense to use the primal simplex optimizer instead.
If this parameter is on and the problem is dualized and furthermore the simplex optimizer
is chosen to be the primal (dual) one, then it is switched to the dual (primal).
Possible values:
Default value:
MSK OFF
sol filter keep basic
MSK IPAR SOL FILTER KEEP BASIC
Description:
If turned on, then basic and super basic constraints and variables are written to the solution
le independent of the lter setting.
Possible values:
Default value:
MSK OFF
sol filter keep ranged
MSK IPAR SOL FILTER KEEP RANGED
Description:
If turned on, then ranged constraints and variables are written to the solution le indepen-
dent of the lter setting.
Possible values:
Default value:
MSK OFF
sol quoted names
MSK IPAR SOL QUOTED NAMES
Description:
If this options is turned on, then MOSEK will quote names that contains blanks while
writing the solution le. Moreover when reading leading and trailing quotes will be stripped
of.
Possible values:
Default value:
MSK OFF
sol read name width
MSK IPAR SOL READ NAME WIDTH
Description:
When a solution is read by MOSEK and some constraint, variable or cone names contain
blanks, then a maximum name width much be specied. A negative value implies that no
name contain blanks.
Possible Values:
Default value:
-1
sol read width
MSK IPAR SOL READ WIDTH
Description:
Controls the maximal acceptable width of line in the solutions when read by MOSEK.
Possible Values:
Any positive number greater than 80.
Default value:
1024
solution callback
MSK IPAR SOLUTION CALLBACK
Description:
Possible values:
Default value:
MSK OFF
timing level
MSK IPAR TIMING LEVEL
Description:
Possible Values:
Any integer greater or equal to 0.
Default value:
1
warning level
MSK IPAR WARNING LEVEL
Description:
Warning level.
Possible Values:
Default value:
1
write bas constraints
MSK IPAR WRITE BAS CONSTRAINTS
Description:
Controls whether the constraint section is written to the basic solution le.
Possible values:
Default value:
MSK ON
write bas head
MSK IPAR WRITE BAS HEAD
Description:
Controls whether the header section is written to the basic solution le.
Possible values:
Default value:
MSK ON
write bas variables
MSK IPAR WRITE BAS VARIABLES
Description:
Controls whether the variables section is written to the basic solution le.
Possible values:
Default value:
MSK ON
write data compressed
MSK IPAR WRITE DATA COMPRESSED
Description:
Controls whether the data le is compressed while it is written. 0 means no compression
while higher values mean more compression.
Possible Values:
Default value:
0
write data format
MSK IPAR WRITE DATA FORMAT
Description:
Controls the le format when writing task data to a le.
Possible values:
MSK DATA FORMAT XML The data le is an XML formatted le.
MSK DATA FORMAT FREE MPS The data data a free MPS formatted le.
MSK DATA FORMAT EXTENSION The le extension is used to determine the data le format.
MSK DATA FORMAT MPS The data le is MPS formatted.
MSK DATA FORMAT LP The data le is LP formatted.
MSK DATA FORMAT MBT The data le is a MOSEK binary task le.
MSK DATA FORMAT OP The data le is an optimization problem formatted le.
Default value:
MSK DATA FORMAT EXTENSION
write data param
MSK IPAR WRITE DATA PARAM
Description:
If this option is turned on the parameter settings are written to the data le as parameters.
Possible values:
Default value:
MSK OFF
write free con
MSK IPAR WRITE FREE CON
Description:
Controls whether the free constraints are written to the data le.
Possible values:
Default value:
MSK OFF
write generic names
MSK IPAR WRITE GENERIC NAMES
Description:
Controls whether the generic names or user-dened names are used in the data le.
Possible values:
Default value:
MSK OFF
write generic names io
MSK IPAR WRITE GENERIC NAMES IO
Description:
Possible Values:
Default value:
1
write int constraints
MSK IPAR WRITE INT CONSTRAINTS
Description:
Controls whether the constraint section is written to the integer solution le.
Possible values:
Default value:
MSK ON
write int head
MSK IPAR WRITE INT HEAD
Description:
Controls whether the header section is written to the integer solution le.
Possible values:
Default value:
MSK ON
write int variables
MSK IPAR WRITE INT VARIABLES
Description:
Controls whether the variables section is written to the integer solution le.
Possible values:
Default value:
MSK ON
write lp line width
MSK IPAR WRITE LP LINE WIDTH
Description:
Maximum width of line in an LP le written by MOSEK.
Possible Values:
Any positive number.
Default value:
80
write lp quoted names
MSK IPAR WRITE LP QUOTED NAMES
Description:
If this option is turned on, then MOSEK will quote invalid LP names when writing an LP
le.
Possible values:
Default value:
MSK ON
write lp strict format
MSK IPAR WRITE LP STRICT FORMAT
Description:
Possible values:
Default value:
MSK OFF
write lp terms per line
MSK IPAR WRITE LP TERMS PER LINE
Description:
Maximum number of terms on a single line in an LP le written by MOSEK. 0 means
unlimited.
Possible Values:
Default value:
10
write mps int
MSK IPAR WRITE MPS INT
Description:
Controls if marker records are written to the MPS le to indicate whether variables are
integer restricted.
Possible values:
Default value:
MSK ON
write mps obj sense
MSK IPAR WRITE MPS OBJ SENSE
Description:
If turned o, the objective sense section is not written to the MPS le.
Possible values:
Default value:
MSK ON
write mps quoted names
MSK IPAR WRITE MPS QUOTED NAMES
Description:
If a name contains spaces (blanks) when writing an MPS le, then the quotes will be
removed.
Possible values:
Default value:
MSK ON
write mps strict
MSK IPAR WRITE MPS STRICT
Description:
Controls whether the written MPS le satises the MPS format strictly or not.
Possible values:
Default value:
MSK OFF
write precision
MSK IPAR WRITE PRECISION
Description:
Controls the precision with which double numbers are printed in the MPS data le. In
general it is not worthwhile to use a value higher than 15.
Possible Values:
Default value:
8
write sol constraints
MSK IPAR WRITE SOL CONSTRAINTS
Description:
Controls whether the constraint section is written to the solution le.
Possible values:
Default value:
MSK ON
write sol head
MSK IPAR WRITE SOL HEAD
Description:
Controls whether the header section is written to the solution le.
Possible values:
Default value:
MSK ON
write sol variables
MSK IPAR WRITE SOL VARIABLES
Description:
Controls whether the variables section is written to the solution le.
Possible values:
Default value:
MSK ON
write task inc sol
MSK IPAR WRITE TASK INC SOL
Description:
Possible values:
Default value:
MSK ON
write xml mode
MSK IPAR WRITE XML MODE
Description:
Controls if linear coecients should be written by row or column when writing in the XML
le format.
Possible values:
MSK WRITE XML MODE COL Write in column order.
MSK WRITE XML MODE ROW Write in row order.
Default value:
MSK WRITE XML MODE ROW
H.4 String parameter types
MSK SPAR DATA FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
MSK SPAR DEBUG FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
MOSEK debug le.
MSK SPAR FEASREPAIR NAME PREFIX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Feasibility repair name prex.
MSK SPAR FEASREPAIR NAME SEPARATOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Feasibility repair name separator.
MSK SPAR FEASREPAIR NAME WSUMVIOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Feasibility repair name violation name.
MSK SPAR PARAM COMMENT SIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Solution le comment character.
MSK SPAR PARAM READ FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
MSK SPAR PARAM WRITE FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
MSK SPAR READ MPS BOU NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Name of the BOUNDS vector used. An empty name means that the rst BOUNDS vector is
used.
MSK SPAR READ MPS OBJ NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Objective name in the MPS le.
MSK SPAR READ MPS RAN NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Name of the RANGE vector used. An empty name means that the rst RANGE vector is used.
MSK SPAR READ MPS RHS NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
MSK SPAR SENSITIVITY FILE NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Sensitivity report le name.
H.4. STRING PARAMETER TYPES 273
MSK SPAR SENSITIVITY RES FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Name of the sensitivity report output le.
Solution le lter.
Solution le lter.
Solution le lter.
Solution le lter.
MSK SPAR STAT FILE NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Statistics le name.
MSK SPAR STAT KEY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
MSK SPAR STAT NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
MSK SPAR WRITE LP GEN VAR NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Added variable names in the LP les.
bas sol file name
MSK SPAR BAS SOL FILE NAME
Description:
Possible Values:
Any valid le name.
Default value:
""
data file name
MSK SPAR DATA FILE NAME
Description:
Possible Values:
Any valid le name.
Default value:
""
debug file name
MSK SPAR DEBUG FILE NAME
Description:
MOSEK debug le.
Possible Values:
Any valid le name.
Default value:
""
feasrepair name prefix
MSK SPAR FEASREPAIR NAME PREFIX
Description:
Not applicable.
Possible Values:
Any valid string.
Default value:
"MSK-"
feasrepair name separator
MSK SPAR FEASREPAIR NAME SEPARATOR
Description:
Not applicable.
Possible Values:
Any valid string.
Default value:
"-"
feasrepair name wsumviol
MSK SPAR FEASREPAIR NAME WSUMVIOL
Description:
The constraint and variable associated with the total weighted sum of violations are each
given the name of this parameter postxed with CON and VAR respectively.
Possible Values:
Any valid string.
Default value:
"WSUMVIOL"
int sol file name
MSK SPAR INT SOL FILE NAME
Description:
Possible Values:
Any valid le name.
Default value:
""
itr sol file name
MSK SPAR ITR SOL FILE NAME
Description:
Possible Values:
Any valid le name.
Default value:
""
param comment sign
MSK SPAR PARAM COMMENT SIGN
Description:
Only the rst character in this string is used. It is considered as a start of comment sign in
the MOSEK parameter le. Spaces are ignored in the string.
Possible Values:
Any valid string.
Default value:
"%%"
param read file name
MSK SPAR PARAM READ FILE NAME
Description:
Possible Values:
Any valid le name.
Default value:
""
param write file name
MSK SPAR PARAM WRITE FILE NAME
Description:
Possible Values:
Any valid le name.
Default value:
""
read mps bou name
MSK SPAR READ MPS BOU NAME
Description:
Name of the BOUNDS vector used. An empty name means that the rst BOUNDS vector
is used.
Possible Values:
Any valid MPS name.
Default value:
""
read mps obj name
MSK SPAR READ MPS OBJ NAME
Description:
Name of the free constraint used as objective function. An empty name means that the
rst constraint is used as objective function.
Possible Values:
Any valid MPS name.
Default value:
""
read mps ran name
MSK SPAR READ MPS RAN NAME
Description:
Name of the RANGE vector used. An empty name means that the rst RANGE vector is
used.
Possible Values:
Any valid MPS name.
Default value:
""
read mps rhs name
MSK SPAR READ MPS RHS NAME
Description:
Possible Values:
Any valid MPS name.
Default value:
""
sensitivity file name
MSK SPAR SENSITIVITY FILE NAME
Description:
Not applicable.
Possible Values:
Any valid string.
Default value:
""
sensitivity res file name
MSK SPAR SENSITIVITY RES FILE NAME
Description:
Not applicable.
Possible Values:
Any valid string.
Default value:
""
sol filter xc low
MSK SPAR SOL FILTER XC LOW
Description:
A lter used to determine which constraints should be listed in the solution le. A value of
0.5 means that all constraints having xc[i]>0.5 should be listed, whereas +0.5 means
that all constraints having xc[i]>=blc[i]+0.5 should be listed. An empty lter means
that no lter is applied.
Possible Values:
Any valid lter.
Default value:
""
sol filter xc upr
MSK SPAR SOL FILTER XC UPR
Description:
A lter used to determine which constraints should be listed in the solution le. A value of
0.5 means that all constraints having xc[i]<0.5 should be listed, whereas -0.5 means
all constraints having xc[i]<=buc[i]-0.5 should be listed. An empty lter means that no
lter is applied.
Possible Values:
Any valid lter.
Default value:
""
sol filter xx low
MSK SPAR SOL FILTER XX LOW
Description:
A lter used to determine which variables should be listed in the solution le. A value of
0.5 means that all constraints having xx[j]>=0.5 should be listed, whereas +0.5 means
that all constraints having xx[j]>=blx[j]+0.5 should be listed. An empty lter means no
lter is applied.
Possible Values:
Any valid lter..
Default value:
""
sol filter xx upr
MSK SPAR SOL FILTER XX UPR
Description:
A lter used to determine which variables should be listed in the solution le. A value of
0.5 means that all constraints having xx[j]<0.5 should be printed, whereas -0.5 means
all constraints having xx[j]<=bux[j]-0.5 should be listed. An empty lter means no lter
is applied.
Possible Values:
Any valid le name.
Default value:
""
stat file name
MSK SPAR STAT FILE NAME
Description:
Statistics le name.
Possible Values:
Any valid le name.
Default value:
""
stat key
MSK SPAR STAT KEY
Description:
Possible Values:
Any valid XML string.
Default value:
""
stat name
MSK SPAR STAT NAME
Description:
Possible Values:
Any valid XML string.
Default value:
""
write lp gen var name
MSK SPAR WRITE LP GEN VAR NAME
Description:
Sometimes when an LP le is written additional variables must be inserted. They will have
the prex denoted by this parameter.
Possible Values:
Any valid string.
Default value:
"xmskgen"
Appendix I
Symbolic constants reference
I.1 Constraint or variable access modes
Value Name
Description
0 MSK ACC VAR
Access data by columns (variable orinted)
1 MSK ACC CON
Access data by rows (constraint oriented)
I.2 Function opcode
Value Name
Description
1 MSK ADOP SUB
Subtract two operands.
4 MSK ADOP POW
First operand to the power the second operand.
7 MSK ADOP RET
Return one operand.
0 MSK ADOP ADD
Add two operands.
5 MSK ADOP EXP
Exponential function of one oparand.
2 MSK ADOP MUL
Multiply two operands.
3 MSK ADOP DIV
Divide two operands.
6 MSK ADOP LOG
continued on next page
281
282 APPENDIX I. SYMBOLIC CONSTANTS REFERENCE
continued from previous page
Logarithm function of one operand.
I.3 Function operand type
Value Name
Description
2 MSK ADOPTYPE VARIABLE
Operand refers to a variable.
0 MSK ADOPTYPE NONE
Operand not used.
1 MSK ADOPTYPE CONSTANT
Operand refers to a constant.
3 MSK ADOPTYPE REFERENCE
Operand refers to the result of another operation.
I.4 Basis identication
Value Name
Description
1 MSK BI ALWAYS
Basis identication is always performed even if the interior-point op-
timizer terminates abnormally.
2 MSK BI NO ERROR
Basis identication is performed if the interior-point optimizer termi-
nates without an error.
0 MSK BI NEVER
Never do basis identication.
3 MSK BI IF FEASIBLE
Basis identication is not performed if the interior-point optimizer
terminates with a problem status saying that the problem is primal
or dual infeasible.
4 MSK BI OTHER
Try another BI method.
I.5 Bound keys
Value Name
Description
2 MSK BK FX
I.7. PROGRESS CALL-BACK CODES 283
The constraint or variable is xed.
0 MSK BK LO
The constraint or variable has a nite lower bound and an innite
upper bound.
3 MSK BK FR
The constraint or variable is free.
1 MSK BK UP
The constraint or variable has an innite lower bound and an nite
upper bound.
4 MSK BK RA
The constraint or variable is ranged.
I.6 Species the branching direction.
Value Name
Description
2 MSK BRANCH DIR DOWN
The mixed-integer optimizer always chooses the down branch rst.
1 MSK BRANCH DIR UP
The mixed-integer optimizer always chooses the up branch rst.
0 MSK BRANCH DIR FREE
The mixed-integer optimizer decides which branch to choose.
I.7 Progress call-back codes
Value Name
Description
44 MSK CALLBACK END INTPNT
The call-back function is called when the interior-point optimizer is
terminated.
21 MSK CALLBACK BEGIN PRIMAL DUAL SIMPLEX BI
The call-back function is called from within the basis identication
procedure when the primal-dual simplex clean-up phase is started.
48 MSK CALLBACK END NETWORK PRIMAL SIMPLEX
The call-back function is called when the primal network simplex
optimizer is terminated.
99 MSK CALLBACK READ ADD CONS
A chunk of constraints has been read from a problem le.
115 MSK CALLBACK UPDATE PRIMAL SIMPLEX BI
procedure at an intermediate point in the primal simplex clean-
up phase. The frequency of the call-backs is controlled by the
MSK IPAR LOG SIM FREQ parameter.
46 MSK CALLBACK END MIO
The call-back function is called when the mixed-integer optimizer is
terminated.
13 MSK CALLBACK BEGIN NETWORK DUAL SIMPLEX
The call-back function is called when the dual network simplex opti-
mizer is started.
35 MSK CALLBACK END CONCURRENT
Concurrent optimizer is terminated.
93 MSK CALLBACK NEW INT MIO
The call-back function is called after a new integer solution has been
located by the mixed-integer optimizer.
88 MSK CALLBACK IM PRIMAL SIMPLEX
The call-back function is called at an intermediate point in the primal
simplex optimizer.
64 MSK CALLBACK END SIMPLEX NETWORK DETECT
The call-back function is called when the network detection procedure
is terminated.
47 MSK CALLBACK END NETWORK DUAL SIMPLEX
The call-back function is called when the dual network simplex opti-
mizer is terminated.
72 MSK CALLBACK IM INTPNT
The call-back function is called at an intermediate stage within the
interior-point optimizer where the information database has not been
updated.
68 MSK CALLBACK IM DUAL BI
procedure at an intermediate point in the dual phase.
8 MSK CALLBACK BEGIN FULL CONVEXITY CHECK
Begin full convexity check.
3 MSK CALLBACK BEGIN DUAL BI
procedure when the dual phase is started.
4 MSK CALLBACK BEGIN DUAL SENSITIVITY
Dual sensitivity analysis is started.
79 MSK CALLBACK IM MIO PRIMAL SIMPLEX
The call-back function is called at an intermediate point in the mixed-
integer optimizer while running the primal simplex optimizer.
19 MSK CALLBACK BEGIN PRIMAL BI
procedure when the primal phase is started.
100 MSK CALLBACK READ ADD QNZ
A chunk of Q non-zeos has been read from a problem le.
1 MSK CALLBACK BEGIN CONCURRENT
Concurrent optimizer is started.
104 MSK CALLBACK UPDATE DUAL BI
procedure at an intermediate point in the dual phase.
70 MSK CALLBACK IM DUAL SIMPLEX
The call-back function is called at an intermediate point in the dual
simplex optimizer.
11 MSK CALLBACK BEGIN LICENSE WAIT
Begin waiting for license.
81 MSK CALLBACK IM NETWORK PRIMAL SIMPLEX
The call-back function is called at an intermediate point in the primal
network simplex optimizer.
49 MSK CALLBACK END NETWORK SIMPLEX
The call-back function is called when the simplex network optimizer
is terminated.
32 MSK CALLBACK CONIC
The call-back function is called from within the conic optimizer after
the information database has been updated.
89 MSK CALLBACK IM QO REFORMULATE
The call-back function is called at an intermediate stage of the QP to
SOCP reformulation.
2 MSK CALLBACK BEGIN CONIC
The call-back function is called when the conic optimizer is started.
106 MSK CALLBACK UPDATE DUAL SIMPLEX BI
The call-back function is called from within the basis identica-
tion procedure at an intermediate point in the dual simplex clean-
51 MSK CALLBACK END OPTIMIZER
The call-back function is called when the optimizer is terminated.
110 MSK CALLBACK UPDATE PRESOLVE
The call-back function is called from within the presolve procedure.
90 MSK CALLBACK IM SIMPLEX
The call-back function is called from within the simplex optimizer at
an intermediate point.
102 MSK CALLBACK READ OPF
The call-back function is called from the OPF reader.
73 MSK CALLBACK IM LICENSE WAIT
MOSEK is waiting for a license.
15 MSK CALLBACK BEGIN NETWORK SIMPLEX
The call-back function is called when the simplex network optimizer
is started.
36 MSK CALLBACK END CONIC
The call-back function is called when the conic optimizer is termi-
nated.
107 MSK CALLBACK UPDATE NETWORK DUAL SIMPLEX
The call-back function is called in the dual network simplex optimizer.
26 MSK CALLBACK BEGIN QCQO REFORMULATE
Begin QCQO reformulation.
38 MSK CALLBACK END DUAL SENSITIVITY
Dual sensitivity analysis is terminated.
59 MSK CALLBACK END PRIMAL SIMPLEX BI
procedure when the primal clean-up phase is terminated.
101 MSK CALLBACK READ ADD VARS
A chunk of variables has been read from a problem le.
103 MSK CALLBACK READ OPF SECTION
A chunk of Q non-zeos has been read from a problem le.
74 MSK CALLBACK IM LU
The call-back function is called from within the LU factorization pro-
cedure at an intermediate point.
41 MSK CALLBACK END DUAL SIMPLEX BI
procedure when the dual clean-up phase is terminated.
45 MSK CALLBACK END LICENSE WAIT
End waiting for license.
84 MSK CALLBACK IM PRESOLVE
The call-back function is called from within the presolve procedure
at an intermediate stage.
5 MSK CALLBACK BEGIN DUAL SETUP BI
The call-back function is called when the dual BI phase is started.
43 MSK CALLBACK END INFEAS ANA
The call-back function is called when the infeasibility analyzer is ter-
minated.
92 MSK CALLBACK INTPNT
The call-back function is called from within the interior-point opti-
mizer after the information database has been updated.
111 MSK CALLBACK UPDATE PRIMAL BI
procedure at an intermediate point in the primal phase.
94 MSK CALLBACK NONCOVEX
The call-back function is called from within the nonconvex optimizer
after the information database has been updated.
113 MSK CALLBACK UPDATE PRIMAL DUAL SIMPLEX BI
The call-back function is called from within the basis identica-
tion procedure at an intermediate point in the primal-dual simplex
clean-up phase. The frequency of the call-backs is controlled by the
109 MSK CALLBACK UPDATE NONCONVEX
nonconvex optimizer where the information database has been up-
dated.
37 MSK CALLBACK END DUAL BI
procedure when the dual phase is terminated.
61 MSK CALLBACK END READ
MOSEK has nished reading a problem le.
98 MSK CALLBACK READ ADD CONES
A chunk of cones has been read from a problem le.
25 MSK CALLBACK BEGIN PRIMAL SIMPLEX BI
procedure when the primal simplex clean-up phase is started.
30 MSK CALLBACK BEGIN SIMPLEX NETWORK DETECT
The call-back function is called when the network detection procedure
is started.
97 MSK CALLBACK READ ADD ANZ
A chunk of A non-zeos has been read from a problem le.
86 MSK CALLBACK IM PRIMAL DUAL SIMPLEX
The call-back function is called at an intermediate point in the primal-
dual simplex optimizer.
114 MSK CALLBACK UPDATE PRIMAL SIMPLEX
The call-back function is called in the primal simplex optimizer.
33 MSK CALLBACK DUAL SIMPLEX
The call-back function is called from within the dual simplex opti-
mizer.
71 MSK CALLBACK IM FULL CONVEXITY CHECK
The call-back function is called at an intermediate stage of the full
convexity check.
95 MSK CALLBACK PRIMAL SIMPLEX
The call-back function is called from within the primal simplex opti-
mizer.
16 MSK CALLBACK BEGIN NONCONVEX
The call-back function is called when the nonconvex optimizer is
started.
91 MSK CALLBACK IM SIMPLEX BI
The call-back function is called from within the basis identi-
cation procedure at an intermediate point in the simplex clean-
6 MSK CALLBACK BEGIN DUAL SIMPLEX
The call-back function is called when the dual simplex optimizer
started.
24 MSK CALLBACK BEGIN PRIMAL SIMPLEX
The call-back function is called when the primal simplex optimizer is
started.
50 MSK CALLBACK END NONCONVEX
The call-back function is called when the nonconvex optimizer is ter-
minated.
23 MSK CALLBACK BEGIN PRIMAL SETUP BI
The call-back function is called when the primal BI setup is started.
17 MSK CALLBACK BEGIN OPTIMIZER
The call-back function is called when the optimizer is started.
27 MSK CALLBACK BEGIN READ
MOSEK has started reading a problem le.
82 MSK CALLBACK IM NONCONVEX
nonconvex optimizer where the information database has not been
updated.
58 MSK CALLBACK END PRIMAL SIMPLEX
The call-back function is called when the primal simplex optimizer is
terminated.
55 MSK CALLBACK END PRIMAL DUAL SIMPLEX BI
procedure when the primal-dual clean-up phase is terminated.
66 MSK CALLBACK IM BI
procedure at an intermediate point.
80 MSK CALLBACK IM NETWORK DUAL SIMPLEX
The call-back function is called at an intermediate point in the dual
network simplex optimizer.
39 MSK CALLBACK END DUAL SETUP BI
The call-back function is called when the dual BI phase is terminated.
34 MSK CALLBACK END BI
The call-back function is called when the basis identication proce-
dure is terminated.
57 MSK CALLBACK END PRIMAL SETUP BI
The call-back function is called when the primal BI setup is termi-
nated.
31 MSK CALLBACK BEGIN WRITE
MOSEK has started writing a problem le.
63 MSK CALLBACK END SIMPLEX BI
procedure when the simplex clean-up phase is terminated.
56 MSK CALLBACK END PRIMAL SENSITIVITY
Primal sensitivity analysis is terminated.
28 MSK CALLBACK BEGIN SIMPLEX
The call-back function is called when the simplex optimizer is started.
52 MSK CALLBACK END PRESOLVE
The call-back function is called when the presolve is completed.
96 MSK CALLBACK QCONE
The call-back function is called from within the Qcone optimizer.
9 MSK CALLBACK BEGIN INFEAS ANA
The call-back function is called when the infeasibility analyzer is
started.
20 MSK CALLBACK BEGIN PRIMAL DUAL SIMPLEX
The call-back function is called when the primal-dual simplex opti-
mizer is started.
22 MSK CALLBACK BEGIN PRIMAL SENSITIVITY
Primal sensitivity analysis is started.
7 MSK CALLBACK BEGIN DUAL SIMPLEX BI
procedure when the dual simplex clean-up phase is started.
60 MSK CALLBACK END QCQO REFORMULATE
End QCQO reformulation.
87 MSK CALLBACK IM PRIMAL SENSIVITY
The call-back function is called at an intermediate stage of the primal
65 MSK CALLBACK END WRITE
MOSEK has nished writing a problem le.
40 MSK CALLBACK END DUAL SIMPLEX
The call-back function is called when the dual simplex optimizer is
terminated.
112 MSK CALLBACK UPDATE PRIMAL DUAL SIMPLEX
The call-back function is called in the primal-dual simplex optimizer.
29 MSK CALLBACK BEGIN SIMPLEX BI
procedure when the simplex clean-up phase is started.
10 MSK CALLBACK BEGIN INTPNT
The call-back function is called when the interior-point optimizer is
started.
69 MSK CALLBACK IM DUAL SENSIVITY
The call-back function is called at an intermediate stage of the dual
62 MSK CALLBACK END SIMPLEX
The call-back function is called when the simplex optimizer is termi-
nated.
53 MSK CALLBACK END PRIMAL BI
procedure when the primal phase is terminated.
75 MSK CALLBACK IM MIO
integer optimizer.
105 MSK CALLBACK UPDATE DUAL SIMPLEX
The call-back function is called in the dual simplex optimizer.
77 MSK CALLBACK IM MIO INTPNT
integer optimizer while running the interior-point optimizer.
54 MSK CALLBACK END PRIMAL DUAL SIMPLEX
The call-back function is called when the primal-dual simplex opti-
mizer is terminated.
67 MSK CALLBACK IM CONIC
conic optimizer where the information database has not been updated.
78 MSK CALLBACK IM MIO PRESOLVE
integer optimizer while running the presolve.
0 MSK CALLBACK BEGIN BI
The basis identication procedure has been started.
76 MSK CALLBACK IM MIO DUAL SIMPLEX
integer optimizer while running the dual simplex optimizer.
116 MSK CALLBACK WRITE OPF
The call-back function is called from the OPF writer.
108 MSK CALLBACK UPDATE NETWORK PRIMAL SIMPLEX
The call-back function is called in the primal network simplex opti-
mizer.
42 MSK CALLBACK END FULL CONVEXITY CHECK
End full convexity check.
83 MSK CALLBACK IM ORDER
The call-back function is called from within the matrix ordering pro-
cedure at an intermediate point.
85 MSK CALLBACK IM PRIMAL BI
procedure at an intermediate point in the primal phase.
18 MSK CALLBACK BEGIN PRESOLVE
The call-back function is called when the presolve is started.
12 MSK CALLBACK BEGIN MIO
I.11. CPU TYPE 291
The call-back function is called when the mixed-integer optimizer is
started.
14 MSK CALLBACK BEGIN NETWORK PRIMAL SIMPLEX
The call-back function is called when the primal network simplex
optimizer is started.
I.8 Types of convexity checks.
Value Name
Description
1 MSK CHECK CONVEXITY SIMPLE
Perform simple and fast convexity check.
0 MSK CHECK CONVEXITY NONE
No convexity check.
2 MSK CHECK CONVEXITY FULL
Perform a full convexity check.
I.9 Compression types
Value Name
Description
2 MSK COMPRESS GZIP
The type of compression used is gzip compatible.
0 MSK COMPRESS NONE
No compression is used.
1 MSK COMPRESS FREE
The type of compression used is chosen automatically.
I.10 Cone types
Value Name
Description
0 MSK CT QUAD
The cone is a quadratic cone.
1 MSK CT RQUAD
The cone is a rotated quadratic cone.
I.11 CPU type
Value Name
Description
8 MSK CPU POWERPC G5
A G5 PowerPC CPU.
9 MSK CPU INTEL PM
An Intel PM cpu.
1 MSK CPU GENERIC
An generic CPU type for the platform
0 MSK CPU UNKNOWN
An unknown CPU.
7 MSK CPU AMD OPTERON
An AMD Opteron (64 bit).
6 MSK CPU INTEL ITANIUM2
An Intel Itanium2.
4 MSK CPU AMD ATHLON
An AMD Athlon.
5 MSK CPU HP PARISC20
An HP PA RISC version 2.0 CPU.
3 MSK CPU INTEL P4
An Intel Pentium P4 or Intel Xeon.
2 MSK CPU INTEL P3
An Intel Pentium P3.
10 MSK CPU INTEL CORE2
An Intel CORE2 cpu.
I.12 Data format types
Value Name
Description
5 MSK DATA FORMAT XML
The data le is an XML formatted le.
6 MSK DATA FORMAT FREE MPS
The data data a free MPS formatted le.
0 MSK DATA FORMAT EXTENSION
The le extension is used to determine the data le format.
1 MSK DATA FORMAT MPS
The data le is MPS formatted.
2 MSK DATA FORMAT LP
The data le is LP formatted.
3 MSK DATA FORMAT MBT
The data le is a MOSEK binary task le.
4 MSK DATA FORMAT OP
The data le is an optimization problem formatted le.
I.13. DOUBLE INFORMATION ITEMS 293
I.13 Double information items
Value Name
Description
13 MSK DINF INTPNT PRIMAL FEAS
Primal feasibility measure reported by the interior-point or Qcone
optimizers. (For the interior-point optimizer this measure does not
directly related to the original problem because a homogeneous model
is employed).
58 MSK DINF SOL ITR MAX PCNI
Maximal primal cone infeasibility in the interior-point solution. Up-
dated at the end of the optimization.
32 MSK DINF RD TIME
Time spent reading the data le.
28 MSK DINF PRESOLVE ELI TIME
Total time spent in the eliminator since the presolve was invoked.
22 MSK DINF MIO OPTIMIZER TIME
Time spent in the optimizer while solving the relaxtions.
10 MSK DINF INTPNT FACTOR NUM FLOPS
An estimate of the number of ops used in the factorization.
25 MSK DINF MIO TIME
Time spent in the mixed-integer optimizer.
4 MSK DINF BI DUAL TIME
Time spent within the dual phase basis identication procedure since
its invocation.
37 MSK DINF SIM NETWORK TIME
Time spent in the network simplex optimizer since invoking it.
30 MSK DINF PRESOLVE TIME
Total time (in seconds) spent in the presolve since it was invoked.
29 MSK DINF PRESOLVE LINDEP TIME
Total time spent in the linear dependency checker since the presolve
was invoked.
33 MSK DINF SIM DUAL TIME
Time spent in the dual simplex optimizer since invoking it.
60 MSK DINF SOL ITR MAX PINTI
Maximal primal integer infeasibility in the interior-point solution.
Updated at the end of the optimization.
38 MSK DINF SIM OBJ
Objective value reported by the simplex optimizer.
21 MSK DINF MIO OBJ REL GAP
Given that the mixed-integer optimizer has computed a feasible so-
lution and a bound on the optimal objective value, then this item
contains the relative gap dened by
[(objective value of feasible solution) (objective bound)[
max(, [(objective value of feasible solution)[)
.
where is given by the paramater MSK DPAR MIO REL GAP CONST. Oth-
erwise it has the value -1.0.
48 MSK DINF SOL BAS PRIMAL OBJ
Primal objective value of the basic solution. Updated at the end of
the optimization.
17 MSK DINF MIO HEURISTIC TIME
Time spent in the optimizer while solving the relaxtions.
57 MSK DINF SOL ITR MAX PBI
Maximal primal bound infeasibility in the interior-point solution. Up-
45 MSK DINF SOL BAS MAX PBI
Maximal primal bound infeasibility in the basic solution. Updated at
the end of the optimization.
27 MSK DINF OPTIMIZER TIME
Total time spent in the optimizer since it was invoked.
55 MSK DINF SOL ITR MAX DCNI
Maximal dual cone infeasibility in the interior-point solution. Up-
24 MSK DINF MIO ROOT PRESOLVE TIME
Time spent in while presolveing the root relaxation.
9 MSK DINF INTPNT DUAL OBJ
Dual objective value reported by the interior-point or Qcone opti-
mizer.
15 MSK DINF INTPNT TIME
Time spent within the interior-point optimizer since its invocation.
16 MSK DINF MIO CONSTRUCT SOLUTION OBJ
If MOSEK has successfully constructed an integer feasible solution,
then this item contains the optimal objective value corresponding to
the feasible solution.
34 MSK DINF SIM FEAS
Feasibility measure reported by the simplex optimizer.
56 MSK DINF SOL ITR MAX DEQI
Maximal dual equality infeasibility in the interior-point solution. Up-
40 MSK DINF SIM PRIMAL TIME
Time spent in the primal simplex optimizer since invoking it.
41 MSK DINF SIM TIME
I.13. DOUBLE INFORMATION ITEMS 295
Time spent in the simplex optimizer since invoking it.
36 MSK DINF SIM NETWORK PRIMAL TIME
Time spent in the primal network simplex optimizer since invoking
it.
47 MSK DINF SOL BAS MAX PINTI
Maximal primal integer infeasibility in the basic solution. Updated
at the end of the optimization.
51 MSK DINF SOL INT MAX PINTI
Maximal primal integer infeasibility in the integer solution. Updated
3 MSK DINF BI CLEAN TIME
Time spent within the clean-up phase of the basis identication pro-
cedure since its invocation.
31 MSK DINF QCQO REFORMULATE TIME
Time spent with QP reformulation.
53 MSK DINF SOL ITR DUAL OBJ
Dual objective value of the interior-point solution. Updated at the
end of the optimization.
49 MSK DINF SOL INT MAX PBI
Maximal primal bound infeasibility in the integer solution. Updated
8 MSK DINF INTPNT DUAL FEAS
Dual feasibility measure reported by the interior-point and Qcone
optimizer. (For the interior-point optimizer this measure does not
directly related to the original problem because a homogeneous model
is employed.)
7 MSK DINF CONCURRENT TIME
Time spent within the concurrent optimizer since its invocation.
11 MSK DINF INTPNT KAP DIV TAU
This measure should converge to zero if the problem has a primal-
dual optimal solution or to innity if problem is (strictly) primal or
dual infeasible. In case the measure is converging towards a positive
but bounded constant the problem is usually ill-posed.
50 MSK DINF SOL INT MAX PEQI
Maximal primal equality infeasibility in the basic solution. Updated
61 MSK DINF SOL ITR PRIMAL OBJ
Primal objective value of the interior-point solution. Updated at the
end of the optimization.
35 MSK DINF SIM NETWORK DUAL TIME
Time spent in the dual network simplex optimizer since invoking it.
20 MSK DINF MIO OBJ INT
The primal objective value corresponding to the best integer feasible
solution. Please note that at least one integer feasible solution must
have located i.e. check MSK IINF MIO NUM INT SOLUTIONS.
26 MSK DINF MIO USER OBJ CUT
If the objective cut is used, then this information item has the value
of the cut.
43 MSK DINF SOL BAS MAX DBI
Maximal dual bound infeasibility in the basic solution. Updated at
46 MSK DINF SOL BAS MAX PEQI
Maximal primal equality infeasibility in the basic solution. Updated
19 MSK DINF MIO OBJ BOUND
The best known bound on the objective function. This value is un-
dened until at least one relaxation has been solved: To see if this is
the case check that MSK IINF MIO NUM RELAX is stricly positive.
0 MSK DINF BI CLEAN DUAL TIME
Time spent within the dual clean-up optimizer of the basis identi-
cation procedure since its invocation.
6 MSK DINF BI TIME
Time spent within the basis identication procedure since its invoca-
tion.
42 MSK DINF SOL BAS DUAL OBJ
Dual objective value of the basic solution. Updated at the end of the
optimization.
1 MSK DINF BI CLEAN PRIMAL DUAL TIME
Time spent within the primal-dual clean-up optimizer of the basis
identication procedure since its invocation.
14 MSK DINF INTPNT PRIMAL OBJ
Primal objective value reported by the interior-point or Qcone opti-
mizer.
12 MSK DINF INTPNT ORDER TIME
Order time (in seconds).
52 MSK DINF SOL INT PRIMAL OBJ
Primal objective value of the integer solution. Updated at the end of
the optimization.
5 MSK DINF BI PRIMAL TIME
Time spent within the primal phase of the basis identication proce-
dure since its invocation.
18 MSK DINF MIO OBJ ABS GAP
I.14. DOUBLE PARAMETERS 297
Given the mixed-integer optimizer has computed a feasible solution
and a bound on the optimal objective value, then this item contains
the absolute gap dened by
[(objective value of feasible solution) (objective bound)[.
Otherwise it has the value -1.0.
44 MSK DINF SOL BAS MAX DEQI
Maximal dual equality infeasibility in the basic solution. Updated at
59 MSK DINF SOL ITR MAX PEQI
Maximal primal equality infeasibility in the interior-point solution.
Updated at the end of the optimization.
39 MSK DINF SIM PRIMAL DUAL TIME
Time spent in the primal-dual simplex optimizer optimizer since in-
voking it.
2 MSK DINF BI CLEAN PRIMAL TIME
Time spent within the primal clean-up optimizer of the basis identi-
cation procedure since its invocation.
54 MSK DINF SOL ITR MAX DBI
Maximal dual bound infeasibility in the interior-point solution. Up-
23 MSK DINF MIO ROOT OPTIMIZER TIME
Time spent in the optimizer while solving the root relaxation.
I.14 Double parameters
Value Name
Description
40 MSK DPAR LOWER OBJ CUT FINITE TRH
If the lower objective cut is less than the value of this parameter value,
then the lower objective cut i.e. MSK DPAR LOWER OBJ CUT is treated
as .
43 MSK DPAR MIO MAX TIME
This parameter limits the maximum time spent by the mixed-integer
optimizer. A negative number means innity.
2 MSK DPAR BASIS TOL S
60 MSK DPAR PRESOLVE TOL S
Absolute zero tolerance employed for s
i
in the presolve.
65 MSK DPAR UPPER OBJ CUT
If either a primal or dual feasible solution is found proving that
the optimal objective value is outside, [MSK DPAR LOWER OBJ CUT,
MSK DPAR UPPER OBJ CUT], then MOSEK is terminated.
16 MSK DPAR INTPNT CO TOL DFEAS
8 MSK DPAR DATA TOL AIJ LARGE
An element in A which is larger than this value in absolute size causes
a warning message to be printed.
49 MSK DPAR MIO TOL ABS GAP
Absolute optimality tolerance employed by the mixed-integer opti-
mizer.
66 MSK DPAR UPPER OBJ CUT FINITE TRH
If the upper objective cut is greater than the value of this value pa-
rameter, then the the upper objective cut MSK DPAR UPPER OBJ CUT
is treated as .
50 MSK DPAR MIO TOL ABS RELAX INT
Absolute relaxation tolerance of the integer constraints. I.e. min([x[
x|, x| [x[) is less than the tolerance then the integer restrictions
assumed to be satised.
0 MSK DPAR ANA SOL INFEAS TOL
If a constraint violates its bound with an amount larger than this
value, the constraint name, index and violation will be printed by the
solution analyzer.
56 MSK DPAR NONCONVEX TOL OPT
55 MSK DPAR NONCONVEX TOL FEAS
42 MSK DPAR MIO HEURISTIC TIME
Minimum amount of time to be used in the heuristic search for a good
feasible integer solution. A negative values implies that the optimizer
decides the amount of time to be spent in the heuristic.
5 MSK DPAR CHECK CONVEXITY REL TOL
This parameter controls when the full convexity check declares a prob-
lem to be non-convex. Increasing this tolerance relaxes the criteria
for declaring the problem non-convex.
A problem is declared non-convex if negative (positive) pivot elements
are detected in the cholesky factor of a matrix which is required to be
PSD (NSD). This parameter controles how much this non-negativity
requirement may be violated.
If d
i
is the pivot element for column i, then the matrix Q is considered
to not be PSD if:
d
i
[Q
ii
[ check convexity rel tol
61 MSK DPAR PRESOLVE TOL X
Absolute zero tolerance employed for x
j
in the presolve.
24 MSK DPAR INTPNT NL TOL MU RED
46 MSK DPAR MIO NEAR TOL REL GAP
The mixed-integer optimizer is terminated when this toler-
ance is satised. This termination criteria is delayed. See
MSK DPAR MIO DISABLE TERM TIME for details.
64 MSK DPAR SIMPLEX ABS TOL PIV
6 MSK DPAR DATA TOL AIJ
Absolute zero tolerance for elements in A. If any value A
ij
is smaller
than this parameter in absolute terms MOSEK will treat the values
as zero and generate a warning.
15 MSK DPAR FEASREPAIR TOL
Tolerance for constraint enforcing upper bound on sum of weighted
violations in feasibility repair.
30 MSK DPAR INTPNT TOL DSAFE
Controls the initial dual starting point used by the interior-point op-
timizer. If the interior-point optimizer converges slowly.
31 MSK DPAR INTPNT TOL INFEAS
Controls when the optimizer declares the model primal or dual infea-
sible. A small number means the optimizer gets more conservative
about declaring the model infeasible.
25 MSK DPAR INTPNT NL TOL NEAR REL
If the MOSEK nonlinear interior-point optimizer cannot compute a
solution that has the prescribed accuracy, then it will multiply the
termination tolerances with value of this parameter. If the solution
then satises the termination criteria, then the solution is denoted
near optimal, near feasible and so forth.
57 MSK DPAR OPTIMIZER MAX TIME
Maximum amount of time the optimizer is allowed to spent on the
optimization. A negative number means innity.
14 MSK DPAR DATA TOL X
Zero tolerance for constraints and variables i.e. if the distance be-
tween the lower and upper bound is less than this value, then the
lower and lower bound is considered identical.
51 MSK DPAR MIO TOL FEAS
Feasibility tolerance for mixed integer solver. Any solution with max-
imum infeasibility below this value will be considered feasible.
47 MSK DPAR MIO REL ADD CUT LIMITED
Controls how many cuts the mixed-integer optimizer is allowed to
add to the problem. Let be the value of this parameter and m the
number constraints, then mixed-integer optimizer is allowed to m
cuts.
32 MSK DPAR INTPNT TOL MU RED
18 MSK DPAR INTPNT CO TOL MU RED
Relative complementarity gap tolerance feasibility tolerance used by
the conic interior-point optimizer.
21 MSK DPAR INTPNT CO TOL REL GAP
Relative gap termination tolerance used by the conic interior-point
optimizer.
39 MSK DPAR LOWER OBJ CUT
If either a primal or dual feasible solution is found prov-
ing that the optimal objective value is outside, the interval
[MSK DPAR LOWER OBJ CUT, MSK DPAR UPPER OBJ CUT], then MOSEK
is terminated.
41 MSK DPAR MIO DISABLE TERM TIME
The termination criteria governed by
is disabled the rst n seconds. This parameter species the number
n. A negative value is identical to innity i.e. the termination criteria
are never checked.
37 MSK DPAR INTPNT TOL REL STEP
Relative step size to the boundary for linear and quadratic optimiza-
tion problems.
54 MSK DPAR MIO TOL X
11 MSK DPAR DATA TOL C HUGE
An element in c which is larger than the value of this parameter in
absolute terms is considered to be huge and generates an error.
59 MSK DPAR PRESOLVE TOL LIN DEP
63 MSK DPAR SIM LU TOL REL PIV
Relative pivot tolerance employed when computing the LU factoriza-
tion of the basis in the simplex optimizers and in the basis identica-
tion procedure.
A value closer to 1.0 generally improves numerical stability but typi-
cally also implies an increase in the computational work.
12 MSK DPAR DATA TOL CJ LARGE
An element in c which is larger than this value in absolute terms
causes a warning message to be printed.
28 MSK DPAR INTPNT NL TOL REL STEP
Relative step size to the boundary for general nonlinear optimization
problems.
38 MSK DPAR INTPNT TOL STEP SIZE
If the step size falls below the value of this parameter, then the
interior-point optimizer assumes that it is stalled. It it does not not
make any progress.
34 MSK DPAR INTPNT TOL PFEAS
Primal feasibility tolerance used for linear and quadratic optimization
problems.
1 MSK DPAR BASIS REL TOL S
Maximum relative dual bound violation allowed in an optimal basic
solution.
17 MSK DPAR INTPNT CO TOL INFEAS
Controls when the conic interior-point optimizer declares the model
primal or dual infeasible. A small number means the optimizer gets
more conservative about declaring the model infeasible.
48 MSK DPAR MIO REL GAP CONST
This value is used to compute the relative gap for the solution to an
integer optimization problem.
58 MSK DPAR PRESOLVE TOL AIJ
Absolute zero tolerance employed for a
ij
in the presolve.
44 MSK DPAR MIO MAX TIME APRX OPT
Number of seconds spent by the mixed-integer optimizer before the
MSK DPAR MIO TOL REL RELAX INT is applied.
33 MSK DPAR INTPNT TOL PATH
Controls how close the interior-point optimizer follows the central
path. A large value of this parameter means the central is followed
very closely. On numerical unstable problems it may be worthwhile
to increase this parameter.
22 MSK DPAR INTPNT NL MERIT BAL
Controls if the complementarity and infeasibility is converging to zero
at about equal rates.
3 MSK DPAR BASIS TOL X
Maximum absolute primal bound violation allowed in an optimal ba-
sic solution.
36 MSK DPAR INTPNT TOL REL GAP
7 MSK DPAR DATA TOL AIJ HUGE
An element in A which is larger than this value in absolute size causes
an error.
10 MSK DPAR DATA TOL BOUND WRN
If a bound value is larger than this value in absolute size, then a
warning message is issued.
9 MSK DPAR DATA TOL BOUND INF
Any bound which in absolute value is greater than this parameter is
considered innite.
35 MSK DPAR INTPNT TOL PSAFE
Controls the initial primal starting point used by the interior-point
optimizer. If the interior-point optimizer converges slowly and/or the
constraint or variable bounds are very large, then it may be worth-
while to increase this value.
19 MSK DPAR INTPNT CO TOL NEAR REL
If MOSEK cannot compute a solution that has the prescribed accu-
racy, then it will multiply the termination tolerances with value of
this parameter. If the solution then satises the termination criteria,
then the solution is denoted near optimal, near feasible and so forth.
4 MSK DPAR CALLBACK FREQ
Controls the time between calls to the progress call-back function.
Hence, if the value of this parameter is for example 10, then the call-
back is called approximately each 10 seconds. A negative value is
equivalent to innity.
In general frequent call-backs may hurt the performance.
26 MSK DPAR INTPNT NL TOL PFEAS
23 MSK DPAR INTPNT NL TOL DFEAS
52 MSK DPAR MIO TOL REL GAP
Relative optimality tolerance employed by the mixed-integer opti-
mizer.
29 MSK DPAR INTPNT TOL DFEAS
Dual feasibility tolerance used for linear and quadratic optimization
problems.
45 MSK DPAR MIO NEAR TOL ABS GAP
I.16. LICENSE FEATURE 303
Relaxed absolute optimality tolerance employed by the mixed-
integer optimizer. This termination criteria is delayed. See
MSK DPAR MIO DISABLE TERM TIME for details.
53 MSK DPAR MIO TOL REL RELAX INT
Relative relaxation tolerance of the integer constraints. I.e (min([x[
x|, x| [x[)) is less than the tolerance times [x[ then the integer
restrictions assumed to be satised.
62 MSK DPAR QCQO REFORMULATE REL DROP TOL
This parameter determines when columns are dropped in incomplete
cholesky factorization doing reformulation of quadratic problems.
13 MSK DPAR DATA TOL QIJ
Absolute zero tolerance for elements in Q matrices.
27 MSK DPAR INTPNT NL TOL REL GAP
20 MSK DPAR INTPNT CO TOL PFEAS
I.15 Feasibility repair types
Value Name
Description
0 MSK FEASREPAIR OPTIMIZE NONE
Do not optimize the feasibility repair problem.
2 MSK FEASREPAIR OPTIMIZE COMBINED
Minimize with original objective subject to minimal weighted viola-
tion of bounds.
1 MSK FEASREPAIR OPTIMIZE PENALTY
Minimize weighted sum of violations.
I.16 License feature
Value Name
Description
2 MSK FEATURE PTOM
Mixed-integer extension.
1 MSK FEATURE PTON
Nonlinear extension.
0 MSK FEATURE PTS
Base system.
3 MSK FEATURE PTOX
Non-convex extension.
I.17 Integer information items.
Value Name
Description
57 MSK IINF RD NUMINTVAR
Number of integer-constrained variables read.
90 MSK IINF SOL BAS SOLSTA
Solution status of the basic solution. Updated after each optimiza-
tion.
97 MSK IINF STO NUM A TRANSPOSES
Number of times the A matrix is transposed. A large number implies
that maxnumanz is too small or an inecient usage of MOSEK. This
will occur in particular if the code alternate between accessing rows
and columns of A.
48 MSK IINF MIO TOTAL NUM OBJ CUTS
Number of obj cuts.
88 MSK IINF SIM SOLVE DUAL
Is non-zero if dual problem is solved.
30 MSK IINF MIO NUMCON
Number of constraints in the problem solved be the mixed-integer
optimizer.
53 MSK IINF OPT NUMVAR
Number of variables in the problem solved when the optimizer is
called
77 MSK IINF SIM NUMVAR
Number of variables in the problem solved by the simplex optimizer.
46 MSK IINF MIO TOTAL NUM LATTICE CUTS
Number of lattice cuts.
71 MSK IINF SIM NETWORK PRIMAL DEG ITER
The number of primal network degenerate iterations.
58 MSK IINF RD NUMQ
Number of nonempty Q matrices read.
0 MSK IINF ANA PRO NUM CON
Number of constraints in the problem.
19 MSK IINF INTPNT FACTOR NUM OFFCOL
Number of columns in the constraint matrix (or Jacobian) that has
an oending structure.
11 MSK IINF ANA PRO NUM VAR INT
Number of general integer variables.
69 MSK IINF SIM NETWORK DUAL INF ITER
The number of iterations taken with dual infeasibility in the network
optimizer.
8 MSK IINF ANA PRO NUM VAR CONT
Number of continuous variables.
I.17. INTEGER INFORMATION ITEMS. 305
65 MSK IINF SIM DUAL ITER
Number of dual simplex iterations during the last optimization.
61 MSK IINF SIM DUAL DEG ITER
The number of dual degenerate iterations.
20 MSK IINF INTPNT ITER
Number of interior-point iterations since invoking the interior-point
optimizer.
45 MSK IINF MIO TOTAL NUM KNAPSUR COVER CUTS
Number of knapsack cover cuts.
33 MSK IINF MIO TOTAL NUM BASIS CUTS
Number of basis cuts.
76 MSK IINF SIM NUMCON
Number of constraints in the problem solved by the simplex optimizer.
83 MSK IINF SIM PRIMAL DUAL ITER
Number of primal dual simplex iterations during the last optimiza-
tion.
5 MSK IINF ANA PRO NUM CON UP
Number of constraints with an upper bound and an innite lower
bound.
36 MSK IINF MIO TOTAL NUM CLIQUE CUTS
Number of clique cuts.
80 MSK IINF SIM PRIMAL DUAL HOTSTART
If 1 then the primal dual simplex algorithm is solving from an ad-
vanced basis.
22 MSK IINF INTPNT SOLVE DUAL
Non-zero if the interior-point optimizer is solving the dual problem.
67 MSK IINF SIM NETWORK DUAL HOTSTART
If 1 then the dual network simplex algorithm is solving from an ad-
vanced basis.
54 MSK IINF OPTIMIZE RESPONSE
The reponse code returned by optimize.
93 MSK IINF SOL ITR PROSTA
Problem status of the interior-point solution. Updated after each
optimization.
60 MSK IINF RD PROTYPE
Problem type.
94 MSK IINF SOL ITR SOLSTA
Solution status of the interior-point solution. Updated after each
optimization.
2 MSK IINF ANA PRO NUM CON FR
Number of unbounded constraints.
81 MSK IINF SIM PRIMAL DUAL HOTSTART LU
If 1 then a valid basis factorization of full rank was located and used
by the primal dual simplex algorithm.
31 MSK IINF MIO NUMINT
Number of integer variables in the problem solved be the mixed-
integer optimizer.
35 MSK IINF MIO TOTAL NUM CARDGUB CUTS
Number of cardgub cuts.
38 MSK IINF MIO TOTAL NUM CONTRA CUTS
Number of contra cuts.
49 MSK IINF MIO TOTAL NUM PLAN LOC CUTS
Number of loc cuts.
64 MSK IINF SIM DUAL INF ITER
The number of iterations taken with dual infeasibility.
32 MSK IINF MIO NUMVAR
Number of variables in the problem solved be the mixed-integer op-
timizer.
27 MSK IINF MIO NUM CUTS
Number of cuts generated by the mixed-integer optimizer.
23 MSK IINF MIO CONSTRUCT SOLUTION
If this item has the value 0, then MOSEK did not try to construct an
initial integer feasible solution. If the item has a positive value, then
MOSEK successfully constructed an initial integer feasible solution.
6 MSK IINF ANA PRO NUM VAR
Number of variables in the problem.
95 MSK IINF STO NUM A CACHE FLUSHES
Number of times the cache of A elements is ushed. A large number
implies that maxnumanz is too small as well as an inecient usage of
MOSEK.
91 MSK IINF SOL INT PROSTA
Problem status of the integer solution. Updated after each optimiza-
tion.
66 MSK IINF SIM NETWORK DUAL DEG ITER
The number of dual network degenerate iterations.
92 MSK IINF SOL INT SOLSTA
Solution status of the integer solution. Updated after each optimiza-
tion.
15 MSK IINF CACHE SIZE L1
L1 cache size used.
16 MSK IINF CACHE SIZE L2
L2 cache size used.
59 MSK IINF RD NUMVAR
Number of variables read.
79 MSK IINF SIM PRIMAL DUAL DEG ITER
The number of degenerate major iterations taken by the primal dual
simplex algorithm.
12 MSK IINF ANA PRO NUM VAR LO
I.17. INTEGER INFORMATION ITEMS. 307
Number of variables with a lower bound and an innite upper bound.
47 MSK IINF MIO TOTAL NUM LIFT CUTS
Number of lift cuts.
89 MSK IINF SOL BAS PROSTA
Problem status of the basic solution. Updated after each optimiza-
tion.
75 MSK IINF SIM NETWORK PRIMAL ITER
Number of primal network simplex iterations during the last opti-
mization.
3 MSK IINF ANA PRO NUM CON LO
Number of constraints with a lower bound and an innite upper
bound.
44 MSK IINF MIO TOTAL NUM GUB COVER CUTS
Number of GUB cover cuts.
68 MSK IINF SIM NETWORK DUAL HOTSTART LU
by the dual network simplex algorithm.
74 MSK IINF SIM NETWORK PRIMAL INF ITER
The number of iterations taken with primal infeasibility in the net-
work optimizer.
84 MSK IINF SIM PRIMAL HOTSTART
If 1 then the primal simplex algorithm is solving from an advanced
basis.
26 MSK IINF MIO NUM BRANCH
Number of branches performed during the optimization.
96 MSK IINF STO NUM A REALLOC
Number of times the storage for storing A has been changed. A large
value may indicates that memory fragmentation may occur.
29 MSK IINF MIO NUM RELAX
Number of relaxations solved during the optimization.
34 MSK IINF MIO TOTAL NUM BRANCH
Number of branches performed during the optimization.
42 MSK IINF MIO TOTAL NUM GCD CUTS
Number of gcd cuts.
41 MSK IINF MIO TOTAL NUM FLOW COVER CUTS
Number of ow cover cuts.
28 MSK IINF MIO NUM INT SOLUTIONS
Number of integer feasible solutions that has been found.
85 MSK IINF SIM PRIMAL HOTSTART LU
by the primal simplex algorithm.
18 MSK IINF CPU TYPE
The type of cpu detected.
1 MSK IINF ANA PRO NUM CON EQ
Number of equality constraints.
13 MSK IINF ANA PRO NUM VAR RA
Number of variables with nite lower and upper bounds.
86 MSK IINF SIM PRIMAL INF ITER
The number of iterations taken with primal infeasibility.
55 MSK IINF RD NUMCON
Number of constraints read.
56 MSK IINF RD NUMCONE
Number of conic constraints read.
10 MSK IINF ANA PRO NUM VAR FR
Number of free variables.
25 MSK IINF MIO NUM ACTIVE NODES
Number of active nodes in the branch and bound tree.
50 MSK IINF MIO TOTAL NUM RELAX
Number of relaxations solved during the optimization.
7 MSK IINF ANA PRO NUM VAR BIN
Number of binary (0-1) variables.
73 MSK IINF SIM NETWORK PRIMAL HOTSTART LU
by the primal network simplex algorithm.
87 MSK IINF SIM PRIMAL ITER
Number of primal simplex iterations during the last optimization.
62 MSK IINF SIM DUAL HOTSTART
If 1 then the dual simplex algorithm is solving from an advanced basis.
24 MSK IINF MIO INITIAL SOLUTION
Is non-zero if an initial integer solution is specied.
21 MSK IINF INTPNT NUM THREADS
Number of threads that the interior-point optimizer is using.
63 MSK IINF SIM DUAL HOTSTART LU
by the dual simplex algorithm.
14 MSK IINF ANA PRO NUM VAR UP
Number of variables with an upper bound and an innite lower bound.
This value is set by
70 MSK IINF SIM NETWORK DUAL ITER
Number of dual network simplex iterations during the last optimiza-
tion.
9 MSK IINF ANA PRO NUM VAR EQ
Number of xed variables.
17 MSK IINF CONCURRENT FASTEST OPTIMIZER
The type of the optimizer that nished rst in a concurrent optimiza-
tion.
51 MSK IINF MIO USER OBJ CUT
If it is non-zero, then the objective cut is used.
I.19. INPUT/OUTPUT MODES 309
43 MSK IINF MIO TOTAL NUM GOMORY CUTS
Number of Gomory cuts.
72 MSK IINF SIM NETWORK PRIMAL HOTSTART
If 1 then the primal network simplex algorithm is solving from an
advanced basis.
40 MSK IINF MIO TOTAL NUM DISAGG CUTS
Number of diasagg cuts.
37 MSK IINF MIO TOTAL NUM COEF REDC CUTS
Number of coef. redc. cuts.
82 MSK IINF SIM PRIMAL DUAL INF ITER
The number of master iterations with dual infeasibility taken by the
primal dual simplex algorithm.
4 MSK IINF ANA PRO NUM CON RA
Number of constraints with nite lower and upper bounds.
39 MSK IINF MIO TOTAL NUM CUTS
Total number of cuts generated by the mixed-integer optimizer.
52 MSK IINF OPT NUMCON
Number of constraints in the problem solved when the optimizer is
called.
78 MSK IINF SIM PRIMAL DEG ITER
The number of primal degenerate iterations.
I.18 Information item types
Value Name
Description
0 MSK INF DOU TYPE
Is a double information type.
2 MSK INF LINT TYPE
Is a long integer.
1 MSK INF INT TYPE
Is an integer.
I.19 Input/output modes
Value Name
Description
0 MSK IOMODE READ
The le is read-only.
1 MSK IOMODE WRITE
The le is write-only. If the le exists then it is truncated when it is
opened. Otherwise it is created when it is opened.
2 MSK IOMODE READWRITE
The le is to read and written.
I.20 Integer parameters
Value Name
Description
175 MSK IPAR SIM STABILITY PRIORITY
125 MSK IPAR READ ADD CONE
Additional number of conic constraints that is made room for in the
problem.
166 MSK IPAR SIM PRIMAL PHASEONE METHOD
204 MSK IPAR WRITE MPS STRICT
Controls whether the written MPS le satises the MPS format
strictly or not.
25 MSK IPAR INFEAS REPORT AUTO
Controls whether an infeasibility report is automatically produced
after the optimization if the problem is primal or dual infeasible.
93 MSK IPAR MIO NODE OPTIMIZER
Controls which optimizer is employed at the non-root nodes in the
mixed-integer optimizer.
118 MSK IPAR PRESOLVE LEVEL
Currently not used.
127 MSK IPAR READ ADD VAR
121 MSK IPAR PRESOLVE USE
Controls whether the presolve is applied to a problem before it is
optimized.
70 MSK IPAR LOG SENSITIVITY OPT
Controls the amount of logging from the optimizers employed during
the sensitivity analysis. 0 means no logging information is produced.
109 MSK IPAR OPF WRITE SOL ITG
If MSK IPAR OPF WRITE SOLUTIONS is MSK ON and an integer solution
is dened, write the integer solution in OPF les.
186 MSK IPAR WRITE BAS HEAD
Controls whether the header section is written to the basic solution
le.
79 MSK IPAR MIO BRANCH PRIORITIES USE
I.20. INTEGER PARAMETERS 311
Controls whether branching priorities are used by the mixed-integer
optimizer.
83 MSK IPAR MIO CUT LEVEL TREE
Controls the cut level employed by the mixed-integer optimizer at
the tree. See MSK IPAR MIO CUT LEVEL ROOT for an explanation of
the parameter values.
188 MSK IPAR WRITE DATA COMPRESSED
Controls whether the data le is compressed while it is written. 0
means no compression while higher values mean more compression.
140 MSK IPAR READ MPS RELAX
If this option is turned on, then mixed integer constraints are ignored
when a problem is read.
106 MSK IPAR OPF WRITE PARAMETERS
129 MSK IPAR READ CON
Expected maximum number of constraints to be read. The option is
only used by fast MPS and LP le readers.
196 MSK IPAR WRITE INT VARIABLES
Controls whether the variables section is written to the integer solu-
tion le.
123 MSK IPAR READ ADD ANZ
Additional number of non-zeros in A that is made room for in the
problem.
36 MSK IPAR INTPNT ORDER METHOD
Controls the ordering strategy used by the interior-point optimizer
when factorizing the Newton equation system.
110 MSK IPAR OPF WRITE SOL ITR
If MSK IPAR OPF WRITE SOLUTIONS is MSK ON and an interior solution
is dened, write the interior solution in OPF les.
69 MSK IPAR LOG SENSITIVITY
Controls the amount of logging during the sensitivity analysis. 0:
Means no logging information is produced. 1: Timing information is
printed. 2: Sensitivity results are printed.
143 MSK IPAR READ QNZ
Expected maximum number of Q non-zeros to be read. The option
is used only by MPS and LP le readers.
59 MSK IPAR LOG INFEAS ANA
Controls amount of output printed by the infeasibility analyzer pro-
cedures. A higher level implies that more information is logged.
168 MSK IPAR SIM PRIMAL SELECTION
Controls the choice of the incoming variable, known as the selection
strategy, in the primal simplex optimizer.
194 MSK IPAR WRITE INT CONSTRAINTS
Controls whether the constraint section is written to the integer so-
lution le.
199 MSK IPAR WRITE LP STRICT FORMAT
148 MSK IPAR SENSITIVITY TYPE
153 MSK IPAR SIM DUAL RESTRICT SELECTION
The dual simplex optimizer can use a so-called restricted selec-
tion/pricing strategy to chooses the outgoing variable. Hence, if
restricted selection is applied, then the dual simplex optimizer rst
choose a subset of all the potential outgoing variables. Next, for some
time it will choose the outgoing variable only among the subset. From
time to time the subset is redened.
A larger value of this parameter implies that the optimizer will be
more aggressive in its restriction strategy, i.e. a value of 0 implies
that the restriction strategy is not applied at all.
62 MSK IPAR LOG MIO FREQ
Controls how frequent the mixed-integer optimizer prints the log line.
It will print line every time MSK IPAR LOG MIO FREQ relaxations have
been solved.
108 MSK IPAR OPF WRITE SOL BAS
If MSK IPAR OPF WRITE SOLUTIONS is MSK ON and a basic solution is
dened, include the basic solution in OPF les.
14 MSK IPAR CHECK TASK DATA
If this feature is turned on, then the task data is checked for bad
values i.e. NaNs. before an optimization is performed.
99 MSK IPAR MIO ROOT OPTIMIZER
Controls which optimizer is employed at the root node in the mixed-
integer optimizer.
191 MSK IPAR WRITE FREE CON
Controls whether the free constraints are written to the data le.
115 MSK IPAR PRESOLVE ELIM FILL
Controls the maximum amount of ll-in that can be created dur-
ing the elimination phase of the presolve. This parameter times
(numcon+numvar) denotes the amount of ll-in.
101 MSK IPAR NONCONVEX MAX ITERATIONS
Maximum number of iterations that can be used by the nonconvex
optimizer.
88 MSK IPAR MIO LOCAL BRANCH NUMBER
192 MSK IPAR WRITE GENERIC NAMES
Controls whether the generic names or user-dened names are used
in the data le.
184 MSK IPAR WARNING LEVEL
Warning level.
51 MSK IPAR LOG BI FREQ
Controls how frequent the optimizer outputs information about the
basis identication and how frequent the user-dened call-back func-
tion is called.
16 MSK IPAR CONCURRENT PRIORITY DUAL SIMPLEX
Priority of the dual simplex algorithm when selecting solvers for con-
current optimization.
67 MSK IPAR LOG PRESOLVE
Controls amount of output printed by the presolve procedure. A
higher level implies that more information is logged.
126 MSK IPAR READ ADD QNZ
Additional number of non-zeros in the Q matrices that is made room
for in the problem.
206 MSK IPAR WRITE SOL CONSTRAINTS
Controls whether the constraint section is written to the solution le.
35 MSK IPAR INTPNT OFF COL TRH
Controls how many oending columns are detected in the Jacobian
of the constraint matrix.
1 means aggressive detection, higher values mean less aggressive de-
tection.
0 means no detection.
128 MSK IPAR READ ANZ
Expected maximum number of A non-zeros to be read. The option
is used only by fast MPS and LP le readers.
92 MSK IPAR MIO MODE
Controls whether the optimizer includes the integer restrictions when
solving a (mixed) integer optimization problem.
134 MSK IPAR READ LP DROP NEW VARS IN BOU
If this option is turned on, MOSEK will drop variables that are de-
ned for the rst time in the bounds section.
71 MSK IPAR LOG SIM
Controls amount of output printed by the simplex optimizer. A higher
level implies that more information is logged.
41 MSK IPAR LIC TRH EXPIRY WRN
If a license feature expires in a numbers days less than the value of
this parameter then a warning will be issued.
182 MSK IPAR SOLUTION CALLBACK
Indicates whether solution call-backs will be performed during the
optimization.
173 MSK IPAR SIM SCALING METHOD
72 MSK IPAR LOG SIM FREQ
Controls how frequent the simplex optimizer outputs information
about the optimization and how frequent the user-dened call-back
function is called.
63 MSK IPAR LOG NONCONVEX
22 MSK IPAR FEASREPAIR OPTIMIZE
198 MSK IPAR WRITE LP QUOTED NAMES
If this option is turned on, then MOSEK will quote invalid LP names
when writing an LP le.
55 MSK IPAR LOG FACTOR
4 MSK IPAR AUTO UPDATE SOL INFO
Controls whether the solution information items are automatically
updated after an optimization is performed.
203 MSK IPAR WRITE MPS QUOTED NAMES
If a name contains spaces (blanks) when writing an MPS le, then
the quotes will be removed.
141 MSK IPAR READ MPS WIDTH
Controls the maximal number of characters allowed in one line of the
MPS le.
183 MSK IPAR TIMING LEVEL
65 MSK IPAR LOG ORDER
82 MSK IPAR MIO CUT LEVEL ROOT
Controls the cut level employed by the mixed-integer optimizer at the
root node. A negative value means a default value determined by the
mixed-integer optimizer is used. By adding the appropriate values
from the following table the employed cut types can be controlled.
GUB cover +2
Flow cover +4
Lifting +8
Plant location +16
Disaggregation +32
Knapsack cover +64
Lattice +128
Gomory +256
Coecient reduction +512
GCD +1024
Obj. integrality +2048
5 MSK IPAR BASIS SOLVE USE PLUS ONE
If a slack variable is in the basis, then the corresponding column in
the basis is a unit vector with -1 in the right position. However, if
this parameter is set to MSK ON, -1 is replaced by 1.
142 MSK IPAR READ Q MODE
94 MSK IPAR MIO NODE SELECTION
Controls the node selection strategy employed by the mixed-integer
optimizer.
2 MSK IPAR ANA SOL PRINT VIOLATED
181 MSK IPAR SOL READ WIDTH
Controls the maximal acceptable width of line in the solutions when
read by MOSEK.
120 MSK IPAR PRESOLVE LINDEP WORK LIM
Is used to limit the amount of work that can done to locate linear
dependencies. In general the higher value this parameter is given the
less work can be used. However, a value of 0 means no limit on the
amount work that can be used.
33 MSK IPAR INTPNT MAX NUM REFINEMENT STEPS
Maximum number of steps to be used by the iterative renement
of the search direction. A negative value implies that the optimizer
Chooses the maximum number of iterative renement steps.
124 MSK IPAR READ ADD CON
Additional number of constraints that is made room for in the prob-
lem.
53 MSK IPAR LOG CONCURRENT
73 MSK IPAR LOG SIM MINOR
159 MSK IPAR SIM MAX ITERATIONS
Maximum number of iterations that can be used by a simplex opti-
mizer.
31 MSK IPAR INTPNT MAX ITERATIONS
Controls the maximum number of iterations allowed in the interior-
point optimizer.
20 MSK IPAR CPU TYPE
Species the CPU type. By default MOSEK tries to auto detect the
CPU type. Therefore, we recommend to change this parameter only
if the auto detection does not work properly.
50 MSK IPAR LOG BI
Controls the amount of output printed by the basis identication
procedure. A higher level implies that more information is logged.
32 MSK IPAR INTPNT MAX NUM COR
Controls the maximum number of correctors allowed by the multiple
corrector procedure. A negative value means that MOSEK is making
the choice.
197 MSK IPAR WRITE LP LINE WIDTH
Maximum width of line in an LP le written by MOSEK.
180 MSK IPAR SOL READ NAME WIDTH
When a solution is read by MOSEK and some constraint, variable or
cone names contain blanks, then a maximum name width much be
specied. A negative value implies that no name contain blanks.
45 MSK IPAR LICENSE DEBUG
This option is used to turn on debugging of the incense manager.
48 MSK IPAR LICENSE WAIT
If all licenses are in use MOSEK returns with an error code. How-
ever, by turning on this parameter MOSEK will wait for an available
license.
1 MSK IPAR ANA SOL BASIS
116 MSK IPAR PRESOLVE ELIMINATOR MAX NUM TRIES
193 MSK IPAR WRITE GENERIC NAMES IO
15 MSK IPAR CONCURRENT NUM OPTIMIZERS
The maximum number of simultaneous optimizations that will be
started by the concurrent optimizer.
169 MSK IPAR SIM REFACTOR FREQ
Controls how frequent the basis is refactorized. The value 0 means
that the optimizer determines the best point of refactorization.
It is strongly recommended NOT to change this parameter.
154 MSK IPAR SIM DUAL SELECTION
Controls the choice of the incoming variable, known as the selection
strategy, in the dual simplex optimizer.
174 MSK IPAR SIM SOLVE FORM
Controls whether the primal or the dual problem is solved by the
primal-/dual- simplex optimizer.
13 MSK IPAR CHECK CONVEXITY
122 MSK IPAR QO SEPARABLE REFORMULATION
Determine if Quadratic programing problems should be reformulated
to separable form.
76 MSK IPAR LP WRITE IGNORE INCOMPATIBLE ITEMS
Controls the result of writing a problem containing incompatible
items to an LP le.
144 MSK IPAR READ TASK IGNORE PARAM
Controls whether MOSEK should ignore the parameter setting de-
ned in the task le and use the default parameter setting instead.
95 MSK IPAR MIO OPTIMIZER MODE
60 MSK IPAR LOG INTPNT
Controls amount of output printed printed by the interior-point op-
timizer. A higher level implies that more information is logged.
61 MSK IPAR LOG MIO
Controls the log level for the mixed-integer optimizer. A higher level
156 MSK IPAR SIM HOTSTART
66 MSK IPAR LOG PARAM
Controls the amount of information printed out about parameter
changes.
189 MSK IPAR WRITE DATA FORMAT
Controls the le format when writing task data to a le.
155 MSK IPAR SIM EXPLOIT DUPVEC
Controls if the simplex optimizers are allowed to exploit duplicated
columns.
78 MSK IPAR MIO BRANCH DIR
Controls whether the mixed-integer optimizer is branching up or down
by default.
29 MSK IPAR INTPNT FACTOR DEBUG LVL
179 MSK IPAR SOL QUOTED NAMES
If this options is turned on, then MOSEK will quote names that
contains blanks while writing the solution le. Moreover when reading
leading and trailing quotes will be stripped of.
46 MSK IPAR LICENSE PAUSE TIME
If MSK IPAR LICENSE WAIT=MSK ON and no license is available, then
MOSEK sleeps a number of milliseconds between each check of
whether a license has become free.
96 MSK IPAR MIO PRESOLVE AGGREGATE
Controls whether the presolve used by the mixed-integer optimizer
tries to aggregate the constraints.
209 MSK IPAR WRITE TASK INC SOL
43 MSK IPAR LICENSE CACHE TIME
Setting this parameter no longer has any eect. Please see
MSK IPAR CACHE LICENSE for an alternative.
9 MSK IPAR BI MAX ITERATIONS
Controls the maximum number of simplex iterations allowed to opti-
mize a basis after the basis identication.
157 MSK IPAR SIM HOTSTART LU
Determines if the simplex optimizer should exploit the initial factor-
ization.
111 MSK IPAR OPF WRITE SOLUTIONS
162 MSK IPAR SIM NETWORK DETECT HOTSTART
This parameter controls has large the network component in rela-
tive terms has to be before it is exploited in a simplex hot-start. The
network component should be equal or larger than
max(MSK_IPAR_SIM_NETWORK_DETECT,MSK_IPAR_SIM_NETWORK_DETECT_HOTSTART)
before it is exploited. If this value is larger than 100 the network ow
component is never detected or exploited.
119 MSK IPAR PRESOLVE LINDEP USE
Controls whether the linear constraints are checked for linear depen-
dencies.
114 MSK IPAR PARAM READ IGN ERROR
104 MSK IPAR OPF WRITE HEADER
81 MSK IPAR MIO CONT SOL
Controls the meaning of the interior-point and basic solutions in
mixed integer problems.
102 MSK IPAR OBJECTIVE SENSE
If the objective sense for the task is undened, then the value of this
parameter is used as the default objective sense.
195 MSK IPAR WRITE INT HEAD
Controls whether the header section is written to the integer solution
le.
40 MSK IPAR INTPNT STARTING POINT
49 MSK IPAR LOG
Controls the amount of log information. The value 0 implies that
all log information is suppressed. A higher level implies that more
Please note that if a task is employed to solve a sequence of optimiza-
tion problems the value of this parameter is reduced by the value
of MSK IPAR LOG CUT SECOND OPT for the second and any subsequent
optimizations.
19 MSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEX
Priority of the primal simplex algorithm when selecting solvers for
concurrent optimization.
138 MSK IPAR READ MPS OBJ SENSE
If turned on, the MPS reader uses the objective sense section. Oth-
erwise the MPS reader ignores it.
10 MSK IPAR CACHE LICENSE
Species if the license is kept checked out for the lifetime of the mosek
environment (on) or returned to the server immediately after the
optimization (o).
Check-in and check-out of licenses have an overhead. Frequent com-
munication with the license server should be avoided.
74 MSK IPAR LOG SIM NETWORK FREQ
Controls how frequent the network simplex optimizer outputs in-
formation about the optimization and how frequent the user-
dened call-back function is called. The network optimizer will
use a logging frequency equal to MSK IPAR LOG SIM FREQ times
MSK IPAR LOG SIM NETWORK FREQ.
28 MSK IPAR INTPNT DIFF STEP
Controls whether dierent step sizes are allowed in the primal and
dual space.
172 MSK IPAR SIM SCALING
Controls how much eort is used in scaling the problem before a
simplex optimizer is used.
200 MSK IPAR WRITE LP TERMS PER LINE
Maximum number of terms on a single line in an LP le written by
MOSEK. 0 means unlimited.
146 MSK IPAR SENSITIVITY ALL
Not applicable.
178 MSK IPAR SOL FILTER KEEP RANGED
If turned on, then ranged constraints and variables are written to the
solution le independent of the lter setting.
7 MSK IPAR BI IGNORE MAX ITER
If the parameter MSK IPAR INTPNT BASIS has the value
MSK BI NO ERROR and the interior-point optimizer has terminated
due to maximum number of iterations, then basis identication is
56 MSK IPAR LOG FEASREPAIR
Controls the amount of output printed when performing feasibility
repair.
39 MSK IPAR INTPNT SOLVE FORM
103 MSK IPAR OPF MAX TERMS PER LINE
The maximum number of terms (linear and quadratic) per line when
an OPF le is written.
205 MSK IPAR WRITE PRECISION
Controls the precision with which double numbers are printed in the
MPS data le. In general it is not worthwhile to use a value higher
than 15.
149 MSK IPAR SIM BASIS FACTOR USE
Controls whether a (LU) factorization of the basis is used in a hot-
start. Forcing a refactorization sometimes improves the stability of
the simplex optimizers, but in most cases there is a performance
penanlty.
210 MSK IPAR WRITE XML MODE
Controls if linear coecients should be written by row or column
when writing in the XML le format.
37 MSK IPAR INTPNT REGULARIZATION USE
6 MSK IPAR BI CLEAN OPTIMIZER
97 MSK IPAR MIO PRESOLVE PROBING
Controls whether the mixed-integer presolve performs probing. Prob-
ing can be very time consuming.
42 MSK IPAR LICENSE ALLOW OVERUSE
24 MSK IPAR INFEAS PREFER PRIMAL
If both certicates of primal and dual infeasibility are supplied then
only the primal is used when this option is turned on.
187 MSK IPAR WRITE BAS VARIABLES
Controls whether the variables section is written to the basic solution
le.
75 MSK IPAR LOG STORAGE
When turned on, MOSEK prints messages regarding the storage usage
and allocation.
98 MSK IPAR MIO PRESOLVE USE
Controls whether presolve is performed by the mixed-integer opti-
mizer.
135 MSK IPAR READ LP QUOTED NAMES
If a name is in quotes when reading an LP le, the quotes will be
removed.
27 MSK IPAR INTPNT BASIS
Controls whether the interior-point optimizer also computes an opti-
mal basis.
54 MSK IPAR LOG CUT SECOND OPT
If a task is employed to solve a sequence of optimization problems,
then the value of the log levels is reduced by the value of this param-
eter. E.g MSK IPAR LOG and MSK IPAR LOG SIM are reduced by the
value of this parameter for the second and any subsequent optimiza-
tions.
137 MSK IPAR READ MPS KEEP INT
Controls whether MOSEK should keep the integer restrictions on the
variables while reading the MPS le.
91 MSK IPAR MIO MAX NUM SOLUTIONS
The mixed-integer optimizer can be terminated after a certain number
of dierent feasible solutions has been located. If this parameter has
the value n and n is strictly positive, then the mixed-integer optimizer
will be terminated when n feasible solutions have been located.
44 MSK IPAR LICENSE CHECK TIME
The parameter species the number of seconds between the checks
of all the active licenses in the MOSEK environment license cache.
These checks are performed to determine if the licenses should be
returned to the server.
208 MSK IPAR WRITE SOL VARIABLES
Controls whether the variables section is written to the solution le.
147 MSK IPAR SENSITIVITY OPTIMIZER
Controls which optimizer is used for optimal partition sensitivity anal-
ysis.
201 MSK IPAR WRITE MPS INT
Controls if marker records are written to the MPS le to indicate
whether variables are integer restricted.
160 MSK IPAR SIM MAX NUM SETBACKS
Controls how many set-backs are allowed within a simplex optimizer.
A set-back is an event where the optimizer moves in the wrong direc-
tion. This is impossible in theory but may happen due to numerical
problems.
21 MSK IPAR DATA CHECK
If this option is turned on, then extensive data checking is enabled.
It will slow down MOSEK but on the other hand help locating bugs.
17 MSK IPAR CONCURRENT PRIORITY FREE SIMPLEX
Priority of the free simplex optimizer when selecting solvers for con-
133 MSK IPAR READ KEEP FREE CON
57 MSK IPAR LOG FILE
If turned on, then some log info is printed when a le is written or
read.
18 MSK IPAR CONCURRENT PRIORITY INTPNT
Priority of the interior-point algorithm when selecting solvers for con-
164 MSK IPAR SIM NON SINGULAR
Controls if the simplex optimizer ensures a non-singular basis, if pos-
sible.
190 MSK IPAR WRITE DATA PARAM
If this option is turned on the parameter settings are written to the
data le as parameters.
150 MSK IPAR SIM DEGEN
105 MSK IPAR OPF WRITE HINTS
Write a hint section with problem dimensions in the beginning of an
OPF le.
117 MSK IPAR PRESOLVE ELIMINATOR USE
Controls whether free or implied free variables are eliminated from
the problem.
0 MSK IPAR ALLOC ADD QNZ
Additional number of Q non-zeros that are allocated space for when
numanz exceeds maxnumqnz during addition of new Q entries.
86 MSK IPAR MIO HOTSTART
136 MSK IPAR READ MPS FORMAT
113 MSK IPAR PARAM READ CASE NAME
139 MSK IPAR READ MPS QUOTED NAMES
If a name is in quotes when reading an MPS le, then the quotes will
be removed.
64 MSK IPAR LOG OPTIMIZER
202 MSK IPAR WRITE MPS OBJ SENSE
If turned o, the objective sense section is not written to the MPS
le.
34 MSK IPAR INTPNT NUM THREADS
Controls the number of threads employed by the interior-point opti-
mizer. If set to a positive number MOSEK will use this number of
threads. If zero the number of threads used will equal the number of
cores detected on the machine.
89 MSK IPAR MIO MAX NUM BRANCHES
Maximum number of branches allowed during the branch and bound
search. A negative value means innite.
165 MSK IPAR SIM PRIMAL CRASH
Controls whether crashing is performed in the primal simplex opti-
mizer.
In general, if a basis consists of more than (100-this parameter
value)% xed variables, then a crash will be performed.
80 MSK IPAR MIO CONSTRUCT SOL
If set to MSK ON and all integer variables have been given a value for
which a feasible mixed integer solution exists, then MOSEK generates
an initial solution to the mixed integer problem by xing all integer
values and solving the remaining problem.
3 MSK IPAR AUTO SORT A BEFORE OPT
Controls whether the elements in each column of A are sorted before
an optimization is performed. This is not required but makes the
optimization more deterministic.
100 MSK IPAR MIO STRONG BRANCH
The value species the depth from the root in which strong branching
is used. A negative value means that the optimizer chooses a default
value automatically.
152 MSK IPAR SIM DUAL PHASEONE METHOD
158 MSK IPAR SIM INTEGER
167 MSK IPAR SIM PRIMAL RESTRICT SELECTION
The primal simplex optimizer can use a so-called restricted selec-
tion/pricing strategy to chooses the outgoing variable. Hence, if re-
stricted selection is applied, then the primal simplex optimizer rst
choose a subset of all the potential incoming variables. Next, for some
time it will choose the incoming variable only among the subset. From
time to time the subset is redened.
A larger value of this parameter implies that the optimizer will be
more aggressive in its restriction strategy, i.e. a value of 0 implies
that the restriction strategy is not applied at all.
130 MSK IPAR READ CONE
Expected maximum number of conic constraints to be read. The
option is used only by fast MPS and LP le readers.
112 MSK IPAR OPTIMIZER
The paramter controls which optimizer is used to optimize the task.
77 MSK IPAR MAX NUM WARNINGS
47 MSK IPAR LICENSE SUPPRESS EXPIRE WRNS
Controls whether license features expire warnings are suppressed.
207 MSK IPAR WRITE SOL HEAD
Controls whether the header section is written to the solution le.
185 MSK IPAR WRITE BAS CONSTRAINTS
Controls whether the constraint section is written to the basic solution
le.
84 MSK IPAR MIO FEASPUMP LEVEL
Feasibility pump is a heuristic designed to compute an initial feasible
solution. A value of 0 implies that the feasibility pump heuristic
is not used. A value of -1 implies that the mixed-integer optimizer
decides how the feasibility pump heuristic is used. A larger value than
1 implies that the feasibility pump is employed more aggressively.
Normally a value beyond 3 is not worthwhile.
23 MSK IPAR INFEAS GENERIC NAMES
Controls whether generic names are used when an infeasible subprob-
lem is created.
161 MSK IPAR SIM NETWORK DETECT
The simplex optimizer is capable of exploiting a network ow com-
ponent in a problem. However it is only worthwhile to exploit the
network ow component if it is suciently large. This parameter
controls how large the network component has to be in relative
terms before it is exploited. For instance a value of 20 means at least
20% of the model should be a network before it is exploited. If this
value is larger than 100 the network ow component is never detected
or exploited.
68 MSK IPAR LOG RESPONSE
Controls amount of output printed when response codes are reported.
A higher level implies that more information is logged.
26 MSK IPAR INFEAS REPORT LEVEL
Controls the amount of information presented in an infeasibility re-
port. Higher values imply more information.
11 MSK IPAR CACHE SIZE L1
Species the size of the cache of the computer. This parameter is
potentially very important for the eciency on computers if MOSEK
cannot determine the cache size automatically. If the cache size is
negative, then MOSEK tries to determine the value automatically.
12 MSK IPAR CACHE SIZE L2
Species the size of the cache of the computer. This parameter is
potentially very important for the eciency on computers where MO-
SEK cannot determine the cache size automatically. If the cache size
is negative, then MOSEK tries to determine the value automatically.
176 MSK IPAR SIM SWITCH OPTIMIZER
The simplex optimizer sometimes chooses to solve the dual problem
instead of the primal problem. This implies that if you have chosen
to use the dual simplex optimizer and the problem is dualized, then
it actually makes sense to use the primal simplex optimizer instead.
If this parameter is on and the problem is dualized and furthermore
the simplex optimizer is chosen to be the primal (dual) one, then it
is switched to the dual (primal).
131 MSK IPAR READ DATA COMPRESSED
If this option is turned on,it is assumed that the data le is com-
pressed.
8 MSK IPAR BI IGNORE NUM ERROR
If the parameter MSK IPAR INTPNT BASIS has the value
MSK BI NO ERROR and the interior-point optimizer has termi-
nated due to a numerical problem, then basis identication is
107 MSK IPAR OPF WRITE PROBLEM
52 MSK IPAR LOG CHECK CONVEXITY
Controls logging in convexity check on quadratic problems. Set to a
positive value to turn logging on.
If a quadratic coecient matrix is found to violate the requirement of
PSD (NSD) then a list of negative (positive) pivot elements is printed.
The absolute value of the pivot elements is also shown.
132 MSK IPAR READ DATA FORMAT
151 MSK IPAR SIM DUAL CRASH
In general if a basis consists of more than (100-this parameter value)%
xed variables, then a crash will be performed.
163 MSK IPAR SIM NETWORK DETECT METHOD
Controls which type of detection method the network extraction
should use.
145 MSK IPAR READ VAR
Expected maximum number of variable to be read. The option is
used only by MPS and LP le readers.
58 MSK IPAR LOG HEAD
170 MSK IPAR SIM REFORMULATION
Controls if the simplex optimizers are allowed to reformulate the prob-
lem.
171 MSK IPAR SIM SAVE LU
Controls if the LU factorization stored should be replaced with the
LU factorization corresponding to the initial basis.
30 MSK IPAR INTPNT FACTOR METHOD
90 MSK IPAR MIO MAX NUM RELAXS
Maximum number of relaxations allowed during the branch and
bound search. A negative value means innite.
177 MSK IPAR SOL FILTER KEEP BASIC
If turned on, then basic and super basic constraints and variables are
written to the solution le independent of the lter setting.
85 MSK IPAR MIO HEURISTIC LEVEL
Controls the heuristic employed by the mixed-integer optimizer to lo-
cate an initial good integer feasible solution. A value of zero means
the heuristic is not used at all. A larger value than 0 means that
a gradually more sophisticated heuristic is used which is computa-
tionally more expensive. A negative value implies that the optimizer
chooses the heuristic. Normally a value around 3 to 5 should be
optimal.
87 MSK IPAR MIO KEEP BASIS
Controls whether the integer presolve keeps bases in memory. This
speeds on the solution process at cost of bigger memory consumption.
38 MSK IPAR INTPNT SCALING
Controls how the problem is scaled before the interior-point optimizer
is used.
I.21 Language selection constants
Value Name
Description
1 MSK LANG DAN
Danish language selection
0 MSK LANG ENG
English language selection
I.22 Long integer information items.
Value Name
Description
6 MSK LIINF BI CLEAN PRIMAL ITER
Number of primal clean iterations performed in the basis identica-
tion.
9 MSK LIINF INTPNT FACTOR NUM NZ
Number of non-zeros in factorization.
10 MSK LIINF MIO INTPNT ITER
Number of interior-point iterations performed by the mixed-integer
optimizer.
4 MSK LIINF BI CLEAN PRIMAL DUAL ITER
Number of primal-dual clean iterations performed in the basis iden-
tication.
3 MSK LIINF BI CLEAN PRIMAL DUAL DEG ITER
Number of primal-dual degenerate clean iterations performed in the
basis identication.
I.24. CONTINUOUS MIXED-INTEGER SOLUTION TYPE 327
2 MSK LIINF BI CLEAN PRIMAL DEG ITER
Number of primal degenerate clean iterations performed in the basis
identication.
1 MSK LIINF BI CLEAN DUAL ITER
Number of dual clean iterations performed in the basis identication.
13 MSK LIINF RD NUMQNZ
Number of Q non-zeros.
12 MSK LIINF RD NUMANZ
Number of non-zeros in A that is read.
8 MSK LIINF BI PRIMAL ITER
Number of primal pivots performed in the basis identication.
7 MSK LIINF BI DUAL ITER
Number of dual pivots performed in the basis identication.
0 MSK LIINF BI CLEAN DUAL DEG ITER
Number of dual degenerate clean iterations performed in the basis
identication.
11 MSK LIINF MIO SIMPLEX ITER
Number of simplex iterations performed by the mixed-integer opti-
mizer.
5 MSK LIINF BI CLEAN PRIMAL DUAL SUB ITER
Number of primal-dual subproblem clean iterations performed in the
basis identication.
I.23 Mark
Value Name
Description
0 MSK MARK LO
The lower bound is selected for sensitivity analysis.
1 MSK MARK UP
The upper bound is selected for sensitivity analysis.
I.24 Continuous mixed-integer solution type
Value Name
Description
2 MSK MIO CONT SOL ITG
The reported interior-point and basic solutions are a solution to the
problem with all integer variables xed at the value they have in the
integer solution. A solution is only reported in case the problem has
a primal feasible solution.
0 MSK MIO CONT SOL NONE
No interior-point or basic solution are reported when the mixed-
integer optimizer is used.
1 MSK MIO CONT SOL ROOT
The reported interior-point and basic solutions are a solution to the
root node problem when mixed-integer optimizer is used.
3 MSK MIO CONT SOL ITG REL
In case the problem is primal feasible then the reported interior-point
and basic solutions are a solution to the problem with all integer
variables xed at the value they have in the integer solution. If the
problem is primal infeasible, then the solution to the root node prob-
lem is reported.
I.25 Integer restrictions
Value Name
Description
0 MSK MIO MODE IGNORED
The integer constraints are ignored and the problem is solved as a
continuous problem.
2 MSK MIO MODE LAZY
Integer restrictions should be satised if an optimizer is available for
the problem.
1 MSK MIO MODE SATISFIED
Integer restrictions should be satised.
I.26 Mixed-integer node selection types
Value Name
Description
5 MSK MIO NODE SELECTION PSEUDO
The optimizer employs selects the node based on a pseudo cost esti-
mate.
4 MSK MIO NODE SELECTION HYBRID
The optimizer employs a hybrid strategy.
0 MSK MIO NODE SELECTION FREE
The optimizer decides the node selection strategy.
3 MSK MIO NODE SELECTION WORST
The optimizer employs a worst bound node selection strategy.
2 MSK MIO NODE SELECTION BEST
The optimizer employs a best bound node selection strategy.
I.29. NETWORK DETECTION METHOD 329
1 MSK MIO NODE SELECTION FIRST
The optimizer employs a depth rst node selection strategy.
I.27 MPS le format type
Value Name
Description
0 MSK MPS FORMAT STRICT
It is assumed that the input le satises the MPS format strictly.
1 MSK MPS FORMAT RELAXED
It is assumed that the input le satises a slightly relaxed version of
the MPS format.
2 MSK MPS FORMAT FREE
It is assumed that the input le satises the free MPS format. This
implies that spaces are not allowed in names. Otherwise the format
is free.
I.28 Message keys
Value Name
Description
1000 MSK MSG READING FILE
None
1001 MSK MSG WRITING FILE
None
1100 MSK MSG MPS SELECTED
None
I.29 Network detection method
Value Name
Description
1 MSK NETWORK DETECT SIMPLE
The network detection should use a very simple heuristic.
2 MSK NETWORK DETECT ADVANCED
The network detection should use a more advanced heuristic.
0 MSK NETWORK DETECT FREE
The network detection is free.
I.30 Objective sense types
Value Name
Description
1 MSK OBJECTIVE SENSE MINIMIZE
The problem should be minimized.
0 MSK OBJECTIVE SENSE UNDEFINED
The objective sense is undened.
2 MSK OBJECTIVE SENSE MAXIMIZE
The problem should be maximized.
I.31 On/o
Value Name
Description
1 MSK ON
Switch the option on.
0 MSK OFF
Switch the option o.
I.32 Optimizer types
Value Name
Description
1 MSK OPTIMIZER INTPNT
The interior-point optimizer is used.
10 MSK OPTIMIZER CONCURRENT
The optimizer for nonconvex nonlinear problems.
8 MSK OPTIMIZER MIXED INT
The mixed-integer optimizer.
5 MSK OPTIMIZER DUAL SIMPLEX
The dual simplex optimizer is used.
0 MSK OPTIMIZER FREE
The optimizer is chosen automatically.
6 MSK OPTIMIZER PRIMAL DUAL SIMPLEX
The primal dual simplex optimizer is used.
2 MSK OPTIMIZER CONIC
The optimizer for problems having conic constraints.
9 MSK OPTIMIZER NONCONVEX
The optimizer for nonconvex nonlinear problems.
3 MSK OPTIMIZER QCONE
I.35. PRESOLVE METHOD. 331
For internal use only.
4 MSK OPTIMIZER PRIMAL SIMPLEX
The primal simplex optimizer is used.
7 MSK OPTIMIZER FREE SIMPLEX
One of the simplex optimizers is used.
I.33 Ordering strategies
Value Name
Description
5 MSK ORDER METHOD NONE
No ordering is used.
2 MSK ORDER METHOD APPMINLOC2
A variant of the approximate minimum local-ll-in ordering is used.
1 MSK ORDER METHOD APPMINLOC1
Approximate minimum local-ll-in ordering is used.
4 MSK ORDER METHOD GRAPHPAR2
An alternative graph partitioning based ordering.
0 MSK ORDER METHOD FREE
The ordering method is chosen automatically.
3 MSK ORDER METHOD GRAPHPAR1
Graph partitioning based ordering.
I.34 Parameter type
Value Name
Description
0 MSK PAR INVALID TYPE
Not a valid parameter.
3 MSK PAR STR TYPE
Is a string parameter.
1 MSK PAR DOU TYPE
Is a double parameter.
2 MSK PAR INT TYPE
Is an integer parameter.
I.35 Presolve method.
Value Name
Description
1 MSK PRESOLVE MODE ON
The problem is presolved before it is optimized.
0 MSK PRESOLVE MODE OFF
The problem is not presolved before it is optimized.
2 MSK PRESOLVE MODE FREE
It is decided automatically whether to presolve before the problem is
optimized.
I.36 Problem data items
Value Name
Description
0 MSK PI VAR
Item is a variable.
2 MSK PI CONE
Item is a cone.
1 MSK PI CON
Item is a constraint.
I.37 Problem types
Value Name
Description
2 MSK PROBTYPE QCQO
The problem is a quadratically constrained optimization problem.
0 MSK PROBTYPE LO
The problem is a linear optimization problem.
4 MSK PROBTYPE CONIC
A conic optimization.
3 MSK PROBTYPE GECO
General convex optimization.
5 MSK PROBTYPE MIXED
General nonlinear constraints and conic constraints. This combina-
tion can not be solved by MOSEK.
1 MSK PROBTYPE QO
The problem is a quadratic optimization problem.
I.38 Problem status keys
I.39. INTERPRETATION OF QUADRATIC TERMS IN MPS FILES 333
Value Name
Description
6 MSK PRO STA PRIM AND DUAL INFEAS
The problem is primal and dual infeasible.
4 MSK PRO STA PRIM INFEAS
The problem is primal infeasible.
7 MSK PRO STA ILL POSED
The problem is ill-posed. For example, it may be primal and dual
feasible but have a positive duality gap.
0 MSK PRO STA UNKNOWN
Unknown problem status.
2 MSK PRO STA PRIM FEAS
The problem is primal feasible.
8 MSK PRO STA NEAR PRIM AND DUAL FEAS
The problem is at least nearly primal and dual feasible.
10 MSK PRO STA NEAR DUAL FEAS
The problem is at least nearly dual feasible.
11 MSK PRO STA PRIM INFEAS OR UNBOUNDED
The problem is either primal infeasible or unbounded. This may occur
for mixed-integer problems.
1 MSK PRO STA PRIM AND DUAL FEAS
The problem is primal and dual feasible.
5 MSK PRO STA DUAL INFEAS
The problem is dual infeasible.
9 MSK PRO STA NEAR PRIM FEAS
The problem is at least nearly primal feasible.
3 MSK PRO STA DUAL FEAS
The problem is dual feasible.
I.39 Interpretation of quadratic terms in MPS les
Value Name
Description
0 MSK Q READ ADD
All elements in a Q matrix are assumed to belong to the lower trian-
gular part. Duplicate elements in a Q matrix are added together.
1 MSK Q READ DROP LOWER
All elements in the strict lower triangular part of the Q matrices are
dropped.
2 MSK Q READ DROP UPPER
All elements in the strict upper triangular part of the Q matrices are
dropped.
I.40 Response codes
Value Name
Description
352 MSK RES WRN SOL FILE IGNORED VAR
One or more lines in the variable section were ignored when reading
a solution le.
1218 MSK RES ERR PARAM TYPE
The parameter type is invalid.
1203 MSK RES ERR INDEX IS TOO SMALL
An index in an argument is too small.
2501 MSK RES ERR INV MARKI
Invalid value in marki.
803 MSK RES WRN PRESOLVE BAD PRECISION
The presolve estimates that the model is specied with insucient
precision.
1500 MSK RES ERR INV PROBLEM
Invalid problem type. Probably a nonconvex problem has been spec-
ied.
1268 MSK RES ERR INV SKX
Invalid value in skx.
2505 MSK RES ERR CANNOT CLONE NL
A task with a nonlinear function call-back cannot be cloned.
1551 MSK RES ERR MIO NO OPTIMIZER
No optimizer is available for the current class of integer optimization
problems.
4009 MSK RES TRM MIO NUM BRANCHES
The mixed-integer optimizer terminated as to the maximum number
of branches was reached.
4004 MSK RES TRM MIO NEAR ABS GAP
The mixed-integer optimizer terminated because the near optimal
absolute gap tolerance was satised.
2001 MSK RES ERR NO DUAL INFEAS CER
A certicate of infeasibility is not available.
1254 MSK RES ERR MUL A ELEMENT
An element in A is dened multiple times.
1170 MSK RES ERR INVALID NAME IN SOL FILE
An invalid name occurred in a solution le.
1114 MSK RES ERR MPS MUL QOBJ
The Q term in the objective is specied multiple times in the MPS
data le.
1063 MSK RES ERR NO INIT ENV
env is not initialized.
1265 MSK RES ERR UNDEF SOLUTION
I.40. RESPONSE CODES 335
MOSEK has the following solution types:
an interior-point solution,
an basic solution,
and an integer solution.
Each optimizer may set one or more of these solutions; e.g by default
a successful optimization with the interior-point optimizer denes the
interior-point solution, and, for linear problems, also the basic solu-
tion. This error occurs when asking for a solution or for information
about a solution that is not dened.
1288 MSK RES ERR LASTJ
Invalid lastj.
1001 MSK RES ERR LICENSE EXPIRED
The license has expired.
3055 MSK RES ERR SEN INDEX INVALID
Invalid range given in the sensitivity le.
1274 MSK RES ERR INV SKN
Invalid value in skn.
1295 MSK RES ERR OBJ Q NOT PSD
The quadratic coecient matrix in the objective is not positive semi-
denite as expected for a minimization problem.
1234 MSK RES ERR INF LINT NAME
A long integer information name is invalid.
903 MSK RES WRN ANA CLOSE BOUNDS
This warning is issued by problem analyzer, if ranged constraints
or variables with very close upper and lower bounds are detected.
One should consider treating such constraints as equalities and such
variables as constants.
1008 MSK RES ERR MISSING LICENSE FILE
MOSEK cannot nd the license le or license server. Usually this
happens if the operating system variable MOSEKLM LICENSE FILE is
not set up appropriately. Please see the MOSEK installation manual
for details.
1235 MSK RES ERR INDEX
An index is out of range.
1350 MSK RES ERR SOL FILE INVALID NUMBER
An invalid number is specied in a solution le.
2800 MSK RES ERR LU MAX NUM TRIES
Could not compute the LU factors of the matrix within the maximum
number of allowed tries.
1267 MSK RES ERR INV SKC
Invalid value in skc.
201 MSK RES WRN DROPPED NZ QOBJ
One or more non-zero elements were dropped in the Q matrix in the
objective.
3000 MSK RES ERR INTERNAL
An internal error occurred. Please report this problem.
1610 MSK RES ERR BASIS FACTOR
The factorization of the basis is invalid.
1204 MSK RES ERR INDEX IS TOO LARGE
An index in an argument is too large.
1154 MSK RES ERR LP INVALID VAR NAME
A variable name is invalid when used in an LP formatted le.
2950 MSK RES ERR NO DUAL FOR ITG SOL
No dual information is available for the integer solution.
1590 MSK RES ERR OVERFLOW
A computation produced an overow i.e. a very large number.
1150 MSK RES ERR LP INCOMPATIBLE
The problem cannot be written to an LP formatted le.
1501 MSK RES ERR MIXED PROBLEM
The problem contains both conic and nonlinear constraints.
1700 MSK RES ERR FEASREPAIR CANNOT RELAX
An optimization problem cannot be relaxed. This is the case e.g. for
general nonlinear optimization problems.
1207 MSK RES ERR PARAM NAME INT
The parameter name is not correct for an integer parameter.
3057 MSK RES ERR SEN SOLUTION STATUS
No optimal solution found to the original problem given for sensitivity
analysis.
1225 MSK RES ERR INF LINT INDEX
A long integer information index is out of range for the specied type.
405 MSK RES WRN TOO MANY BASIS VARS
A basis with too many variables has been specied.
1081 MSK RES ERR SPACE NO INFO
No available information about the space usage.
1205 MSK RES ERR PARAM NAME
The parameter name is not correct.
1106 MSK RES ERR MPS UNDEF VAR NAME
An undened variable name occurred in an MPS le.
200 MSK RES WRN NZ IN UPR TRI
Non-zero elements specied in the upper triangle of a matrix were
ignored.
505 MSK RES WRN LICENSE FEATURE EXPIRE
The license expires.
1263 MSK RES ERR NEGATIVE SURPLUS
Negative surplus.
1404 MSK RES ERR INV QCON SUBK
Invalid value in qcsubk.
1406 MSK RES ERR INV QCON SUBJ
Invalid value in qcsubj.
705 MSK RES WRN ZEROS IN SPARSE ROW
One or more (near) zero elements are specied in a sparse row of
a matrix. It is redundant to specify zero elements. Hence it may
indicate an error.
1198 MSK RES ERR ARGUMENT TYPE
Incorrect argument type.
1017 MSK RES ERR LICENSE MOSEKLM DAEMON
The MOSEKLM license manager daemon is not up and running.
2901 MSK RES ERR INVALID WCHAR
An invalid wchar string is encountered.
1059 MSK RES ERR END OF FILE
End of le reached.
1462 MSK RES ERR NAN IN BUC
u
c
contains an invalid oating point value, i.e. a NaN.
1290 MSK RES ERR NONLINEAR EQUALITY
The model contains a nonlinear equality which denes a nonconvex
set.
1055 MSK RES ERR DATA FILE EXT
The data le format cannot be determined from the le name.
1210 MSK RES ERR PARAM INDEX
Parameter index is out of range.
1285 MSK RES ERR FIRSTI
Invalid firsti.
1000 MSK RES ERR LICENSE
Invalid license.
1299 MSK RES ERR ARGUMENT PERM ARRAY
An invalid permutation array is specied.
85 MSK RES WRN LP DROP VARIABLE
Ignored a variable because the variable was not previously dened.
Usually this implies that a variable appears in the bound section but
not in the objective or the constraints.
1287 MSK RES ERR FIRSTJ
Invalid firstj.
1432 MSK RES ERR USER NLO FUNC
The user-dened nonlinear function reported an error.
1219 MSK RES ERR INF DOU INDEX
A double information index is out of range for the specied type.
1286 MSK RES ERR LASTI
Invalid lasti.
1431 MSK RES ERR USER FUNC RET DATA
An user function returned invalid data.
3900 MSK RES ERR SIZE LICENSE NUMCORES
The computer contains more cpu cores than the license allows for.
1199 MSK RES ERR NR ARGUMENTS
Incorrect number of function arguments.
1293 MSK RES ERR CON Q NOT PSD
The quadratic constraint matrix is not positive semi-denite as ex-
pected for a constraint with nite upper bound. This results in a
nonconvex problem.
63 MSK RES WRN ZERO AIJ
One or more zero elements are specied in A.
2504 MSK RES ERR INV NUMJ
Invalid numj.
1650 MSK RES ERR FACTOR
An error occurred while factorizing a matrix.
3201 MSK RES ERR INVALID BRANCH PRIORITY
An invalid branching priority is specied. It should be nonnegative.
1216 MSK RES ERR PARAM IS TOO SMALL
The parameter value is too small.
1163 MSK RES ERR LP WRITE CONIC PROBLEM
The problem contains cones that cannot be written to an LP format-
ted le.
1002 MSK RES ERR LICENSE VERSION
The license is valid for another version of MOSEK.
1240 MSK RES ERR MAXNUMCON
The maximum number of constraints specied is smaller than the
number of constraints in the task.
1050 MSK RES ERR UNKNOWN
Unknown error.
1162 MSK RES ERR READ LP NONEXISTING NAME
A variable never occurred in objective or constraints.
2503 MSK RES ERR INV NUMI
Invalid numi.
1292 MSK RES ERR NONLINEAR RANGED
The model contains a nonlinear ranged constraint which by denition
denes a nonconvex set.
1047 MSK RES ERR THREAD MUTEX UNLOCK
Could not unlock a mutex.
1100 MSK RES ERR MPS FILE
An error occurred while reading an MPS le.
1156 MSK RES ERR WRITE OPF INVALID VAR NAME
Empty variable names cannot be written to OPF les.
1152 MSK RES ERR LP DUP SLACK NAME
The name of the slack variable added to a ranged constraint already
exists.
2000 MSK RES ERR NO PRIMAL INFEAS CER
A certicate of primal infeasibility is not available.
1158 MSK RES ERR WRITE LP FORMAT
Problem cannot be written as an LP le.
1461 MSK RES ERR NAN IN BLC
l
c
3058 MSK RES ERR SEN NUMERICAL
Numerical diculties encountered performing the sensitivity analysis.
3052 MSK RES ERR SEN INDEX RANGE
Index out of range in the sensitivity analysis le.
1027 MSK RES ERR LICENSE NO SERVER SUPPORT
The license server does not support the requested feature. Possible
reasons for this error include:
The feature has expired.
The features start date is later than todays date.
The version requested is higher than features the highest sup-
ported version.
A corrupted license le.
Try restarting the license and inspect the license server debug le,
usually called lmgrd.log.
66 MSK RES WRN SPAR MAX LEN
A value for a string parameter is longer than the buer that is sup-
posed to hold it.
3050 MSK RES ERR SEN FORMAT
Syntax error in sensitivity analysis le.
1407 MSK RES ERR INV QCON VAL
Invalid value in qcval.
1206 MSK RES ERR PARAM NAME DOU
The parameter name is not correct for a double parameter.
1172 MSK RES ERR OPF PREMATURE EOF
Premature end of le in an OPF le.
1300 MSK RES ERR CONE INDEX
An index of a non-existing cone has been specied.
1470 MSK RES ERR NAN IN C
c contains an invalid oating point value, i.e. a NaN.
1066 MSK RES ERR LIVING TASKS
All tasks associated with an enviroment must be deleted before the
environment is deleted. There are still some undeleted tasks.
1304 MSK RES ERR MAXNUMCONE
The value specied for maxnumcone is too small.
1103 MSK RES ERR MPS NULL CON NAME
An empty constraint name is used in an MPS le.
1417 MSK RES ERR QCON UPPER TRIANGLE
An element in the upper triangle of a Q
k
is specied. Only elements
in the lower triangle should be specied.
1171 MSK RES ERR LP INVALID CON NAME
A constraint name is invalid when used in an LP formatted le.
1125 MSK RES ERR MPS TAB IN FIELD2
A tab char occurred in eld 2.
270 MSK RES WRN MIO INFEASIBLE FINAL
The nal mixed-integer problem with all the integer variables xed
at their optimal values is infeasible.
710 MSK RES WRN ZEROS IN SPARSE COL
One or more (near) zero elements are specied in a sparse column of
a matrix. It is redundant to specify zero elements. Hence, it may
indicate an error.
1433 MSK RES ERR USER NLO EVAL
The user-dened nonlinear function reported an error.
1232 MSK RES ERR INF TYPE
The information type is invalid.
800 MSK RES WRN INCOMPLETE LINEAR DEPENDENCY CHECK
The linear dependency check(s) was not completed and therefore the
A matrix may contain linear dependencies.
503 MSK RES WRN USING GENERIC NAMES
The le writer reverts to generic names because a name is blank.
1056 MSK RES ERR INVALID FILE NAME
An invalid le name has been specied.
804 MSK RES WRN WRITE DISCARDED CFIX
The xed objective term could not be converted to a variable and was
discarded in the output le.
1415 MSK RES ERR QOBJ UPPER TRIANGLE
An element in the upper triangle of Q
o
is specied. Only elements in
the lower triangle should be specied.
1054 MSK RES ERR FILE WRITE
File write error.
1048 MSK RES ERR THREAD CREATE
Could not create a thread. This error may occur if a large number of
environments are created and not deleted again. In any case it is a
good practice to minimize the number of environments created.
1243 MSK RES ERR MAXNUMQNZ
The maximum number of non-zeros specied for the Q matrices is
smaller than the number of non-zeros in the current Q matrices.
2506 MSK RES ERR CANNOT HANDLE NL
A function cannot handle a task with nonlinear function call-backs.
1600 MSK RES ERR NO BASIS SOL
No basic solution is dened.
1131 MSK RES ERR ORD INVALID
Invalid content in branch ordering le.
1303 MSK RES ERR CONE REP VAR
A variable is included multiple times in the cone.
1075 MSK RES ERR INVALID OBJ NAME
An invalid objective name is specied.
1052 MSK RES ERR FILE OPEN
Error while opening a le.
250 MSK RES WRN IGNORE INTEGER
Ignored integer constraints.
1296 MSK RES ERR OBJ Q NOT NSD
The quadratic coecient matrix in the objective is not negative semi-
denite as expected for a maximization problem.
1064 MSK RES ERR INVALID TASK
The task is invalid.
1065 MSK RES ERR NULL POINTER
An argument to a function is unexpectedly a NULL pointer.
3102 MSK RES ERR AD INVALID CODELIST
The code list data was invalid.
3005 MSK RES ERR API FATAL ERROR
An internal error occurred in the API. Please report this problem.
1550 MSK RES ERR INV OPTIMIZER
An invalid optimizer has been chosen for the problem. This means
that the simplex or the conic optimizer is chosen to optimize a non-
linear problem.
1310 MSK RES ERR REMOVE CONE VARIABLE
A variable cannot be removed because it will make a cone invalid.
62 MSK RES WRN LARGE AIJ
A numerically large value is specied for an a
i,j
element in A. The pa-
rameter MSK DPAR DATA TOL AIJ LARGE controls when an a
i,j
is con-
sidered large.
1208 MSK RES ERR PARAM NAME STR
The parameter name is not correct for a string parameter.
1018 MSK RES ERR LICENSE FEATURE
A requested feature is not available in the license le(s). Most likely
due to an incorrect license system setup.
251 MSK RES WRN NO GLOBAL OPTIMIZER
No global optimizer is available.
1040 MSK RES ERR LINK FILE DLL
A le cannot be linked to a stream in the DLL version.
1701 MSK RES ERR FEASREPAIR SOLVING RELAXED
The relaxed problem could not be solved to optimality. Please consult
the log le for further details.
1221 MSK RES ERR INDEX ARR IS TOO SMALL
An index in an array argument is too small.
1259 MSK RES ERR SOLVER PROBTYPE
Problem type does not match the chosen optimizer.
1220 MSK RES ERR INF INT INDEX
An integer information index is out of range for the specied type.
1053 MSK RES ERR FILE READ
File read error.
1440 MSK RES ERR USER NLO EVAL HESSUBI
The user-dened nonlinear function reported an invalid subscript in
the Hessian.
1441 MSK RES ERR USER NLO EVAL HESSUBJ
The user-dened nonlinear function reported an invalid subscript in
the Hessian.
300 MSK RES WRN SOL FILTER
Invalid solution lter is specied.
3100 MSK RES ERR UNB STEP SIZE
A step size in an optimizer was unexpectedly unbounded. For in-
stance, if the step-size becomes unbounded in phase 1 of the simplex
algorithm then an error occurs. Normally this will happen only if the
problem is badly formulated. Please contact MOSEK support if this
error occurs.
4030 MSK RES TRM INTERNAL
The optimizer terminated due to some internal reason. Please contact
MOSEK support.
1110 MSK RES ERR MPS NO OBJECTIVE
No objective is dened in an MPS le.
1403 MSK RES ERR INV QOBJ VAL
Invalid value in qoval.
1400 MSK RES ERR INFINITE BOUND
A numerically huge bound value is specied.
1030 MSK RES ERR OPEN DL
A dynamic link library could not be opened.
3001 MSK RES ERR API ARRAY TOO SMALL
An input array was too short.
1046 MSK RES ERR THREAD MUTEX LOCK
Could not lock a mutex.
1262 MSK RES ERR LAST
Invalid index last. A given index was out of expected range.
1151 MSK RES ERR LP EMPTY
The problem cannot be written to an LP formatted le.
1011 MSK RES ERR SIZE LICENSE VAR
The problem has too many variables to be solved with the available
license.
1062 MSK RES ERR INVALID STREAM
An invalid stream is referenced.
2520 MSK RES ERR INVALID ACCMODE
An invalid access mode is specied.
1250 MSK RES ERR NUMCONLIM
Maximum number of constraints limit is exceeded.
2550 MSK RES ERR MBT INCOMPATIBLE
The MBT le is incompatible with this platform. This results from
reading a le on a 32 bit platform generated on a 64 bit platform.
1104 MSK RES ERR MPS NULL VAR NAME
An empty variable name is used in an MPS le.
72 MSK RES WRN MPS SPLIT BOU VECTOR
A BOUNDS vector is split into several nonadjacent parts in an MPS
le.
1026 MSK RES ERR LICENSE SERVER VERSION
The version specied in the checkout request is greater than the high-
est version number the daemon supports.
1025 MSK RES ERR LICENSE INVALID HOSTID
The host ID specied in the license le does not match the host ID
of the computer.
1045 MSK RES ERR THREAD MUTEX INIT
Could not initialize a mutex.
54 MSK RES WRN LARGE CON FX
An equality constraint is xed to a numerically large value. This can
cause numerical problems.
1280 MSK RES ERR INV NAME ITEM
An invalid name item code is used.
3106 MSK RES ERR AD MISSING RETURN
The code list data was invalid. Missing return operation in function.
53 MSK RES WRN LARGE UP BOUND
A numerically large upper bound value is specied.
3910 MSK RES ERR INFEAS UNDEFINED
The requested value is not dened for this solution type.
901 MSK RES WRN ANA C ZERO
This warning is issued by the problem analyzer, if the coecients in
the linear part of the objective are all zero.
1112 MSK RES ERR MPS MUL CON NAME
A constraint name was specied multiple times in the ROWS section.
1801 MSK RES ERR INVALID IOMODE
Invalid io mode.
1115 MSK RES ERR MPS INV SEC ORDER
The sections in the MPS data le are not in the correct order.
1016 MSK RES ERR LICENSE MAX
Maximum number of licenses is reached.
4007 MSK RES TRM USER CALLBACK
The optimizer terminated due to the return of the user-dened call-
back function.
805 MSK RES WRN CONSTRUCT SOLUTION INFEAS
After xing the integer variables at the suggested values then the
problem is infeasible.
1058 MSK RES ERR INVALID MBT FILE
A MOSEK binary task le is invalid.
1294 MSK RES ERR CON Q NOT NSD
The quadratic constraint matrix is not negative semi-denite as ex-
pected for a constraint with nite lower bound. This results in a
nonconvex problem.
3600 MSK RES ERR XML INVALID PROBLEM TYPE
The problem type is not supported by the XML format.
1231 MSK RES ERR INF INT NAME
An integer information name is invalid.
1107 MSK RES ERR MPS INV CON KEY
An invalid constraint key occurred in an MPS le.
1425 MSK RES ERR FIXED BOUND VALUES
A xed constraint/variable has been specied using the bound keys
but the numerical value of the lower and upper bound is dierent.
4025 MSK RES TRM NUMERICAL PROBLEM
The optimizer terminated due to numerical problems.
3056 MSK RES ERR SEN INVALID REGEXP
Syntax error in regexp or regexp longer than 1024.
52 MSK RES WRN LARGE LO BOUND
A numerically large lower bound value is specied.
3999 MSK RES ERR API INTERNAL
An internal fatal error occurred in an interface function.
70 MSK RES WRN MPS SPLIT RHS VECTOR
An RHS vector is split into several nonadjacent parts in an MPS le.
3053 MSK RES ERR SEN BOUND INVALID UP
Analysis of upper bound requested for an index, where no upper
bound exists.
1702 MSK RES ERR FEASREPAIR INCONSISTENT BOUND
The upper bound is less than the lower bound for a variable or a
constraint. Please correct this before running the feasibility repair.
1449 MSK RES ERR Y IS UNDEFINED
The solution item y is undened.
3200 MSK RES ERR INVALID BRANCH DIRECTION
An invalid branching direction is specied.
1430 MSK RES ERR USER FUNC RET
An user function reported an error.
1750 MSK RES ERR NAME MAX LEN
A name is longer than the buer that is supposed to hold it.
1305 MSK RES ERR CONE TYPE
Invalid cone type specied.
4008 MSK RES TRM MIO NUM RELAXS
The mixed-integer optimizer terminated as the maximum number of
relaxations was reached.
1256 MSK RES ERR INV BKC
Invalid bound key is specied for a constraint.
4020 MSK RES TRM MAX NUM SETBACKS
The optimizer terminated as the maximum number of set-backs was
reached. This indicates numerical problems and a possibly badly
formulated problem.
4015 MSK RES TRM NUM MAX NUM INT SOLUTIONS
The mixed-integer optimizer terminated as the maximum number of
feasible solutions was reached.
3101 MSK RES ERR IDENTICAL TASKS
Some tasks related to this function call were identical. Unique tasks
were expected.
1020 MSK RES ERR LICENSE CANNOT ALLOCATE
The license system cannot allocate the memory required.
904 MSK RES WRN ANA ALMOST INT BOUNDS
This warning is issued by the problem analyzer if a constraint is bound
nearly integral.
1402 MSK RES ERR INV QOBJ SUBJ
Invalid value in qosubj.
1302 MSK RES ERR CONE OVERLAP
A new cone which variables overlap with an existing cone has been
specied.
807 MSK RES WRN CONSTRUCT INVALID SOL ITG
The intial value for one or more of the integer variables is not feasible.
1401 MSK RES ERR INV QOBJ SUBI
Invalid value in qosubi.
1153 MSK RES ERR WRITE MPS INVALID NAME
An invalid name is created while writing an MPS le. Usually this
will make the MPS le unreadable.
1553 MSK RES ERR MIO NOT LOADED
The mixed-integer optimizer is not loaded.
1061 MSK RES ERR NULL TASK
task is a NULL pointer.
1070 MSK RES ERR BLANK NAME
An all blank name has been specied.
3059 MSK RES ERR CONCURRENT OPTIMIZER
An unsupported optimizer was chosen for use with the concurrent
optimizer.
1252 MSK RES ERR TOO SMALL MAXNUMANZ
The maximum number of non-zeros specied for A is smaller than
the number of non-zeros in the current A.
1197 MSK RES ERR ARGUMENT LENNEQ
Incorrect length of arguments.
500 MSK RES WRN LICENSE EXPIRE
The license expires.
1200 MSK RES ERR IN ARGUMENT
A function argument is incorrect.
1051 MSK RES ERR SPACE
Out of space.
1241 MSK RES ERR MAXNUMVAR
The maximum number of variables specied is smaller than the num-
ber of variables in the task.
1800 MSK RES ERR INVALID COMPRESSION
Invalid compression type.
1101 MSK RES ERR MPS INV FIELD
A eld in the MPS le is invalid. Probably it is too wide.
1060 MSK RES ERR NULL ENV
env is a NULL pointer.
3500 MSK RES ERR INTERNAL TEST FAILED
An internal unit test function failed.
501 MSK RES WRN LICENSE SERVER
The license server is not responding.
1122 MSK RES ERR MPS INVALID OBJSENSE
An invalid objective sense is specied.
1168 MSK RES ERR OPF FORMAT
Syntax error in an OPF le
900 MSK RES WRN ANA LARGE BOUNDS
This warning is issued by the problem analyzer, if one or more con-
straint or variable bounds are very large. One should consider omit-
ting these bounds entirely by setting them to +inf or -inf.
1071 MSK RES ERR DUP NAME
The same name was used multiple times for the same problem item
type.
1116 MSK RES ERR MPS MUL CSEC
Multiple CSECTIONs are given the same name.
51 MSK RES WRN LARGE BOUND
A numerically large bound value is specied.
50 MSK RES WRN OPEN PARAM FILE
The parameter le could not be opened.
1291 MSK RES ERR NONCONVEX
The optimization problem is nonconvex.
1615 MSK RES ERR BASIS SINGULAR
The basis is singular and hence cannot be factored.
1155 MSK RES ERR LP FREE CONSTRAINT
Free constraints cannot be written in LP le format.
1445 MSK RES ERR INVALID OBJECTIVE SENSE
An invalid objective sense is specied.
0 MSK RES OK
No error occurred.
3002 MSK RES ERR API CB CONNECT
Failed to connect a callback object.
1253 MSK RES ERR INV APTRE
aptre[j] is strictly smaller than aptrb[j] for some j.
1013 MSK RES ERR OPTIMIZER LICENSE
The optimizer required is not licensed.
1007 MSK RES ERR FILE LICENSE
Invalid license le.
1160 MSK RES ERR LP FORMAT
Syntax error in an LP le.
1237 MSK RES ERR SOLITEM
The solution item number solitem is invalid. Please note that
MSK SOL ITEM SNX is invalid for the basic solution.
1010 MSK RES ERR SIZE LICENSE CON
The problem has too many constraints to be solved with the available
license.
1118 MSK RES ERR MPS CONE OVERLAP
A variable is specied to be a member of several cones.
1090 MSK RES ERR READ FORMAT
The specied format cannot be read.
1408 MSK RES ERR QCON SUBI TOO SMALL
Invalid value in qcsubi.
1580 MSK RES ERR POSTSOLVE
An error occurred during the postsolve. Please contact MOSEK sup-
port.
1215 MSK RES ERR PARAM IS TOO LARGE
The parameter value is too large.
1164 MSK RES ERR LP WRITE GECO PROBLEM
The problem contains general convex terms that cannot be written
to an LP formatted le.
1281 MSK RES ERR PRO ITEM
An invalid problem is used.
1057 MSK RES ERR INVALID SOL FILE NAME
An invalid le name has been specied.
1271 MSK RES ERR INV CONE TYPE STR
Invalid cone type string encountered.
1283 MSK RES ERR INVALID FORMAT TYPE
Invalid format type.
57 MSK RES WRN LARGE CJ
A numerically large value is specied for one c
j
.
1035 MSK RES ERR OLDER DLL
The dynamic link library is older than the specied version.
1019 MSK RES ERR PLATFORM NOT LICENSED
A requested license feature is not available for the required platform.
1119 MSK RES ERR MPS CONE REPEAT
A variable is repeated within the CSECTION.
3051 MSK RES ERR SEN UNDEF NAME
An undened name was encountered in the sensitivity analysis le.
1380 MSK RES ERR HUGE AIJ
A numerically huge value is specied for an a
i,j
element in A. The
parameter MSK DPAR DATA TOL AIJ HUGE controls when an a
i,j
is con-
sidered huge.
71 MSK RES WRN MPS SPLIT RAN VECTOR
A RANGE vector is split into several nonadjacent parts in an MPS
le.
3054 MSK RES ERR SEN BOUND INVALID LO
Analysis of lower bound requested for an index, where no lower bound
exists.
3105 MSK RES ERR AD MISSING OPERAND
The code list data was invalid. Missing operand for operator.
1111 MSK RES ERR MPS SPLITTED VAR
All elements in a column of the A matrix must be specied consecu-
tively. Hence, it is illegal to specify non-zero elements in A for variable
1, then for variable 2 and then variable 1 again.
1080 MSK RES ERR SPACE LEAKING
MOSEK is leaking memory. This can be due to either an incorrect
use of MOSEK or a bug.
1201 MSK RES ERR ARGUMENT DIMENSION
A function argument is of incorrect dimension.
1159 MSK RES ERR READ LP MISSING END TAG
Missing End tag in LP le.
4001 MSK RES TRM MAX TIME
The optimizer terminated at the maximum amount of time.
810 MSK RES WRN CONSTRUCT NO SOL ITG
The construct solution requires an integer solution.
3700 MSK RES ERR INVALID AMPL STUB
Invalid AMPL stub.
1260 MSK RES ERR OBJECTIVE RANGE
Empty objective range.
1238 MSK RES ERR WHICHITEM NOT ALLOWED
whichitem is unacceptable.
1471 MSK RES ERR NAN IN BLX
l
x
1236 MSK RES ERR WHICHSOL
The solution dened by compwhichsol does not exists.
801 MSK RES WRN ELIMINATOR SPACE
The eliminator is skipped at least once due to lack of space.
1049 MSK RES ERR THREAD COND INIT
Could not initialize a condition.
1269 MSK RES ERR INV SK STR
Invalid status key string encountered.
1036 MSK RES ERR NEWER DLL
The dynamic link library is newer than the specied version.
1251 MSK RES ERR NUMVARLIM
Maximum number of variables limit is exceeded.
1113 MSK RES ERR MPS MUL QSEC
Multiple QSECTIONs are specied for a constraint in the MPS data
le.
502 MSK RES WRN EMPTY NAME
A variable or constraint name is empty. The output le may be
invalid.
4003 MSK RES TRM MIO NEAR REL GAP
The mixed-integer optimizer terminated because the near optimal
relative gap tolerance was satised.
80 MSK RES WRN LP OLD QUAD FORMAT
Missing /2 after quadratic expressions in bound or objective.
1272 MSK RES ERR INV CONE TYPE
Invalid cone type code is encountered.
1102 MSK RES ERR MPS INV MARKER
An invalid marker has been specied in the MPS le.
1230 MSK RES ERR INF DOU NAME
A double information name is invalid.
1264 MSK RES ERR NEGATIVE APPEND
Cannot append a negative number.
1270 MSK RES ERR INV SK
Invalid status key code.
1006 MSK RES ERR PROB LICENSE
The software is not licensed to solve the problem.
3104 MSK RES ERR AD INVALID OPERAND
The code list data was invalid. An unknown operand was used.
1015 MSK RES ERR LICENSE SERVER
The license server is not responding.
4005 MSK RES TRM USER BREAK
The optimizer terminated due to a user break.
400 MSK RES WRN TOO FEW BASIS VARS
An incomplete basis has been specied. Too few basis variables are
specied.
1161 MSK RES ERR WRITE LP NON UNIQUE NAME
An auto-generated name is not unique.
1108 MSK RES ERR MPS INV BOUND KEY
An invalid bound key occurred in an MPS le.
1472 MSK RES ERR NAN IN BUX
u
x
1450 MSK RES ERR NAN IN DOUBLE DATA
An invalid oating point value was used in some double data.
1109 MSK RES ERR MPS INV SEC NAME
An invalid section name occurred in an MPS le.
1266 MSK RES ERR BASIS
An invalid basis is specied. Either too many or too few basis vari-
ables are specied.
1257 MSK RES ERR INV BKX
An invalid bound key is specied for a variable.
351 MSK RES WRN SOL FILE IGNORED CON
One or more lines in the constraint section were ignored when reading
a solution le.
4006 MSK RES TRM STALL
The optimizer terminated due to slow progress. Normally there are
three possible reasons for this: Either a bug in MOSEK, the problem
is badly formulated, or, for nonlinear problems, the nonlinear call-
back functions are incorrect.
The solution returned may or may not be of acceptable quality.
Therefore, the solution status should be examined to determine the
status of the solution.
In particular, if a linear optimization problem is solved with the
interior-point optimizer with basis identication turned on, the re-
turned solution may be of acceptable quality, even in the optimizer
stalled.
902 MSK RES WRN ANA EMPTY COLS
This warning is issued by the problem analyzer, if columns, in which
all coecients are zero, are found.
1128 MSK RES ERR MPS INVALID OBJ NAME
An invalid objective name is specied.
1217 MSK RES ERR PARAM VALUE STR
The parameter value string is incorrect.
1222 MSK RES ERR INDEX ARR IS TOO LARGE
An index in an array argument is too large.
1306 MSK RES ERR CONE TYPE STR
Invalid cone type specied.
1405 MSK RES ERR INV QCON SUBI
1760 MSK RES ERR NAME IS NULL
The name buer is a NULL pointer.
1258 MSK RES ERR INV VAR TYPE
An invalid variable type is specied for a variable.
1157 MSK RES ERR LP FILE FORMAT
Syntax error in an LP le.
1021 MSK RES ERR LICENSE CANNOT CONNECT
MOSEK cannot connect to the license server. Most likely the license
server is not up and running.
4002 MSK RES TRM OBJECTIVE RANGE
The optimizer terminated on the bound of the objective range.
350 MSK RES WRN UNDEF SOL FILE NAME
Undened name occurred in a solution.
1255 MSK RES ERR INV BK
Invalid bound key.
1169 MSK RES ERR OPF NEW VARIABLE
Introducing new variables is now allowed. When a [variables] sec-
tion is present, it is not allowed to introduce new variables later in
the problem.
1014 MSK RES ERR FLEXLM
The FLEXlm license manager reported an error.
1275 MSK RES ERR INVALID SURPLUS
Invalid surplus.
65 MSK RES WRN NAME MAX LEN
A name is longer than the buer that is supposed to hold it.
1301 MSK RES ERR CONE SIZE
A cone with too few members is specied.
1261 MSK RES ERR FIRST
Invalid first.
1473 MSK RES ERR NAN IN AIJ
a
i,j
4031 MSK RES TRM INTERNAL STOP
The optimizer terminated for internal reasons. Please contact MO-
SEK support.
1117 MSK RES ERR MPS CONE TYPE
Invalid cone type specied in a CSECTION.
1005 MSK RES ERR SIZE LICENSE
The problem is bigger than the license.
1409 MSK RES ERR QCON SUBI TOO LARGE
1375 MSK RES ERR HUGE C
A huge value in absolute size is specied for one c
j
.
1446 MSK RES ERR UNDEFINED OBJECTIVE SENSE
The objective sense has not been specied before the optimization.
4000 MSK RES TRM MAX ITERATIONS
The optimizer terminated at the maximum number of iterations.
802 MSK RES WRN PRESOLVE OUTOFSPACE
The presolve is incomplete due to lack of space.
1130 MSK RES ERR ORD INVALID BRANCH DIR
An invalid branch direction key is specied.
3103 MSK RES ERR AD INVALID OPERATOR
The code list data was invalid. An unknown operator was used.
1166 MSK RES ERR WRITING FILE
An error occurred while writing le
2502 MSK RES ERR INV MARKJ
Invalid value in markj.
2500 MSK RES ERR NO SOLUTION IN CALLBACK
The required solution is not available.
2900 MSK RES ERR INVALID UTF8
An invalid UTF8 string is encountered.
1105 MSK RES ERR MPS UNDEF CON NAME
An undened constraint name occurred in an MPS le.
1012 MSK RES ERR SIZE LICENSE INTVAR
The problem contains too many integer variables to be solved with
the available license.
3800 MSK RES ERR INT64 TO INT32 CAST
An 32 bit integer could not cast to a 64 bit integer.
1552 MSK RES ERR NO OPTIMIZER VAR TYPE
No optimizer is available for this class of optimization problems.
I.41 Response code type
Value Name
Description
1 MSK RESPONSE WRN
The response code is a warning.
2 MSK RESPONSE TRM
The response code is an optimizer termination status.
4 MSK RESPONSE UNK
I.45. DEGENERACY STRATEGIES 353
The response code does not belong to any class.
0 MSK RESPONSE OK
The response code is OK.
3 MSK RESPONSE ERR
The response code is an error.
I.42 Scaling type
Value Name
Description
0 MSK SCALING METHOD POW2
Scales only with power of 2 leaving the mantissa untouched.
1 MSK SCALING METHOD FREE
The optimizer chooses the scaling heuristic.
I.43 Scaling type
Value Name
Description
1 MSK SCALING NONE
No scaling is performed.
2 MSK SCALING MODERATE
A conservative scaling is performed.
3 MSK SCALING AGGRESSIVE
A very aggressive scaling is performed.
0 MSK SCALING FREE
The optimizer chooses the scaling heuristic.
I.44 Sensitivity types
Value Name
Description
1 MSK SENSITIVITY TYPE OPTIMAL PARTITION
Optimal partition sensitivity analysis is performed.
0 MSK SENSITIVITY TYPE BASIS
Basis sensitivity analysis is performed.
I.45 Degeneracy strategies
Value Name
Description
0 MSK SIM DEGEN NONE
The simplex optimizer should use no degeneration strategy.
3 MSK SIM DEGEN MODERATE
The simplex optimizer should use a moderate degeneration strategy.
4 MSK SIM DEGEN MINIMUM
The simplex optimizer should use a minimum degeneration strategy.
2 MSK SIM DEGEN AGGRESSIVE
The simplex optimizer should use an aggressive degeneration strategy.
1 MSK SIM DEGEN FREE
The simplex optimizer chooses the degeneration strategy.
I.46 Exploit duplicate columns.
Value Name
Description
1 MSK SIM EXPLOIT DUPVEC ON
Allow the simplex optimizer to exploit duplicated columns.
0 MSK SIM EXPLOIT DUPVEC OFF
Disallow the simplex optimizer to exploit duplicated columns.
2 MSK SIM EXPLOIT DUPVEC FREE
The simplex optimizer can choose freely.
I.47 Hot-start type employed by the simplex optimizer
Value Name
Description
0 MSK SIM HOTSTART NONE
The simplex optimizer performs a coldstart.
2 MSK SIM HOTSTART STATUS KEYS
Only the status keys of the constraints and variables are used to
choose the type of hot-start.
1 MSK SIM HOTSTART FREE
The simplex optimize chooses the hot-start type.
I.48 Problem reformulation.
Value Name
Description
1 MSK SIM REFORMULATION ON
I.50. SOLUTION ITEMS 355
Allow the simplex optimizer to reformulate the problem.
3 MSK SIM REFORMULATION AGGRESSIVE
The simplex optimizer should use an aggressive reformulation strat-
egy.
0 MSK SIM REFORMULATION OFF
Disallow the simplex optimizer to reformulate the problem.
2 MSK SIM REFORMULATION FREE
The simplex optimizer can choose freely.
I.49 Simplex selection strategy
Value Name
Description
1 MSK SIM SELECTION FULL
The optimizer uses full pricing.
5 MSK SIM SELECTION PARTIAL
The optimizer uses a partial selection approach. The approach is
usually benecial if the number of variables is much larger than the
number of constraints.
0 MSK SIM SELECTION FREE
The optimizer chooses the pricing strategy.
2 MSK SIM SELECTION ASE
The optimizer uses approximate steepest-edge pricing.
3 MSK SIM SELECTION DEVEX
The optimizer uses devex steepest-edge pricing (or if it is not available
4 MSK SIM SELECTION SE
The optimizer uses steepest-edge selection (or if it is not available an
approximate steep-edge selection).
I.50 Solution items
Value Name
Description
4 MSK SOL ITEM SUC
Lagrange multipliers for upper bounds on the constraints.
0 MSK SOL ITEM XC
Solution for the constraints.
1 MSK SOL ITEM XX
Variable solution.
2 MSK SOL ITEM Y
Lagrange multipliers for equations.
5 MSK SOL ITEM SLX
Lagrange multipliers for lower bounds on the variables.
6 MSK SOL ITEM SUX
Lagrange multipliers for upper bounds on the variables.
7 MSK SOL ITEM SNX
Lagrange multipliers corresponding to the conic constraints on the
variables.
3 MSK SOL ITEM SLC
Lagrange multipliers for lower bounds on the constraints.
I.51 Solution status keys
Value Name
Description
6 MSK SOL STA DUAL INFEAS CER
The solution is a certicate of dual infeasibility.
5 MSK SOL STA PRIM INFEAS CER
The solution is a certicate of primal infeasibility.
0 MSK SOL STA UNKNOWN
Status of the solution is unknown.
8 MSK SOL STA NEAR OPTIMAL
The solution is nearly optimal.
12 MSK SOL STA NEAR PRIM INFEAS CER
The solution is almost a certicate of primal infeasibility.
2 MSK SOL STA PRIM FEAS
The solution is primal feasible.
15 MSK SOL STA NEAR INTEGER OPTIMAL
The primal solution is near integer optimal.
10 MSK SOL STA NEAR DUAL FEAS
The solution is nearly dual feasible.
14 MSK SOL STA INTEGER OPTIMAL
The primal solution is integer optimal.
13 MSK SOL STA NEAR DUAL INFEAS CER
The solution is almost a certicate of dual infeasibility.
11 MSK SOL STA NEAR PRIM AND DUAL FEAS
The solution is nearly both primal and dual feasible.
1 MSK SOL STA OPTIMAL
The solution is optimal.
4 MSK SOL STA PRIM AND DUAL FEAS
The solution is both primal and dual feasible.
9 MSK SOL STA NEAR PRIM FEAS
The solution is nearly primal feasible.
I.54. STRING PARAMETER TYPES 357
3 MSK SOL STA DUAL FEAS
The solution is dual feasible.
I.52 Solution types
Value Name
Description
2 MSK SOL ITG
The integer solution.
0 MSK SOL ITR
The interior solution.
1 MSK SOL BAS
The basic solution.
I.53 Solve primal or dual form
Value Name
Description
1 MSK SOLVE PRIMAL
The optimizer should solve the primal problem.
2 MSK SOLVE DUAL
The optimizer should solve the dual problem.
0 MSK SOLVE FREE
The optimizer is free to solve either the primal or the dual problem.
I.54 String parameter types
Value Name
Description
8 MSK SPAR PARAM COMMENT SIGN
Only the rst character in this string is used. It is considered as
a start of comment sign in the MOSEK parameter le. Spaces are
ignored in the string.
3 MSK SPAR FEASREPAIR NAME PREFIX
Not applicable.
0 MSK SPAR BAS SOL FILE NAME
12 MSK SPAR READ MPS OBJ NAME
Name of the free constraint used as objective function. An empty
name means that the rst constraint is used as objective function.
5 MSK SPAR FEASREPAIR NAME WSUMVIOL
The constraint and variable associated with the total weighted sum
of violations are each given the name of this parameter postxed with
CON and VAR respectively.
4 MSK SPAR FEASREPAIR NAME SEPARATOR
Not applicable.
10 MSK SPAR PARAM WRITE FILE NAME
6 MSK SPAR INT SOL FILE NAME
14 MSK SPAR READ MPS RHS NAME
Name of the RHS used. An empty name means that the rst RHS
vector is used.
21 MSK SPAR STAT FILE NAME
Statistics le name.
24 MSK SPAR WRITE LP GEN VAR NAME
Sometimes when an LP le is written additional variables must be
inserted. They will have the prex denoted by this parameter.
1 MSK SPAR DATA FILE NAME
13 MSK SPAR READ MPS RAN NAME
Name of the RANGE vector used. An empty name means that the
rst RANGE vector is used.
17 MSK SPAR SOL FILTER XC LOW
A lter used to determine which constraints should be listed in the
solution le. A value of 0.5 means that all constraints having
xc[i]>0.5 should be listed, whereas +0.5 means that all con-
straints having xc[i]>=blc[i]+0.5 should be listed. An empty lter
means that no lter is applied.
18 MSK SPAR SOL FILTER XC UPR
A lter used to determine which constraints should be listed in the
xc[i]<0.5 should be listed, whereas -0.5 means all constraints hav-
ing xc[i]<=buc[i]-0.5 should be listed. An empty lter means that
no lter is applied.
11 MSK SPAR READ MPS BOU NAME
Name of the BOUNDS vector used. An empty name means that the
rst BOUNDS vector is used.
20 MSK SPAR SOL FILTER XX UPR
I.55. STATUS KEYS 359
A lter used to determine which variables should be listed in the
xx[j]<0.5 should be printed, whereas -0.5 means all constraints
having xx[j]<=bux[j]-0.5 should be listed. An empty lter means
no lter is applied.
23 MSK SPAR STAT NAME
9 MSK SPAR PARAM READ FILE NAME
7 MSK SPAR ITR SOL FILE NAME
15 MSK SPAR SENSITIVITY FILE NAME
Not applicable.
2 MSK SPAR DEBUG FILE NAME
MOSEK debug le.
22 MSK SPAR STAT KEY
16 MSK SPAR SENSITIVITY RES FILE NAME
Not applicable.
19 MSK SPAR SOL FILTER XX LOW
A lter used to determine which variables should be listed in the
xx[j]>=0.5 should be listed, whereas +0.5 means that all con-
straints having xx[j]>=blx[j]+0.5 should be listed. An empty lter
means no lter is applied.
I.55 Status keys
Value Name
Description
2 MSK SK SUPBAS
The constraint or variable is super basic.
1 MSK SK BAS
The constraint or variable is in the basis.
5 MSK SK FIX
The constraint or variable is xed.
3 MSK SK LOW
The constraint or variable is at its lower bound.
6 MSK SK INF
The constraint or variable is infeasible in the bounds.
0 MSK SK UNK
The status for the constraint or variable is unknown.
4 MSK SK UPR
The constraint or variable is at its upper bound.
I.56 Starting point types
Value Name
Description
1 MSK STARTING POINT GUESS
The optimizer guesses a starting point.
3 MSK STARTING POINT SATISFY BOUNDS
The starting point is choosen to satisfy all the simple bounds on non-
linear variables. If this starting point is employed, then more care
than usual should employed when choosing the bounds on the non-
linear variables. In particular very tight bounds should be avoided.
2 MSK STARTING POINT CONSTANT
The optimizer constructs a starting point by assigning a constant
value to all primal and dual variables. This starting point is normally
robust.
0 MSK STARTING POINT FREE
The starting point is chosen automatically.
I.57 Stream types
Value Name
Description
1 MSK STREAM MSG
Message stream. Log information relating to performance and
progress of the optimization is written to this stream.
3 MSK STREAM WRN
Warning stream. Warning messages are written to this stream.
0 MSK STREAM LOG
Log stream. Contains the aggregated contents of all other streams.
This means that a message written to any other stream will also be
written to this stream.
2 MSK STREAM ERR
Error stream. Error messages are written to this stream.
I.58 Integer values
I.60. XML WRITER OUTPUT MODE 361
Value Name
Description
1024 MSK MAX STR LEN
Maximum string length allowed in MOSEK.
20 MSK LICENSE BUFFER LENGTH
The length of a license key buer.
I.59 Variable types
Value Name
Description
1 MSK VAR TYPE INT
Is an integer variable.
0 MSK VAR TYPE CONT
Is a continuous variable.
I.60 XML writer output mode
Value Name
Description
1 MSK WRITE XML MODE COL
Write in column order.
0 MSK WRITE XML MODE ROW
Write in row order.
Appendix J
Problem analyzer examples
This appendix presents a few examples of the output produced by the problem analyzer described in
Section 12.1. The rst two problems are taken from the MIPLIB 2003 collection, http://miplib.
zib.de/.
J.1 air04
fixed : all ranged : all bin : all
-------------------------------------------------------------------------------
Objective, min cx
range: min |c|: 31.0000 max |c|: 2258.00
distrib: |c| vars
[31, 100) 176
[100, 1e+03) 8084
[1e+03, 2.26e+03] 644
-------------------------------------------------------------------------------
Row nonzeros, A_i
2 2 0.24 0.24
[3, 7] 4 0.49 0.73
[8, 15] 19 2.31 3.04
[16, 31] 80 9.72 12.76
[32, 63] 236 28.68 41.43
[64, 127] 289 35.12 76.55
[128, 255] 186 22.60 99.15
363
364 APPENDIX J. PROBLEM ANALYZER EXAMPLES
[256, 368] 7 0.85 100.00
2 118 1.33 1.33
[3, 7] 2853 32.04 33.37
[8, 15] 5933 66.63 100.00
A nonzeros, A(ij)
range: all |A(ij)| = 1.00000
-------------------------------------------------------------------------------
[1, 10] 823 823
0 8904
[1, 10] 8904
-------------------------------------------------------------------------------
J.2 arki001
lower bd: 82 lower bd: 38 cont: 850
upper bd: 946 fixed : 353 bin : 415
fixed : 20 free : 1 int : 123
ranged : 996
-------------------------------------------------------------------------------
Objective, min cx
range: all |c| in {0.00000, 1.00000}
distrib: |c| vars
0 1387
1 1
-------------------------------------------------------------------------------
Row nonzeros, A_i
1 29 2.77 2.77
2 476 45.42 48.19
[3, 7] 49 4.68 52.86
[8, 15] 56 5.34 58.21
J.3. PROBLEM WITH BOTH LINEAR AND QUADRATIC CONSTRAINTS 365
[16, 31] 64 6.11 64.31
[32, 63] 373 35.59 99.90
[1024, 1046] 1 0.10 100.00
1 381 27.45 27.45
2 19 1.37 28.82
[3, 7] 38 2.74 31.56
[8, 15] 233 16.79 48.34
[16, 29] 717 51.66 100.00
A nonzeros, A(ij)
range: min |A(ij)|: 0.000200000 max |A(ij)|: 2.33067e+07
[0.0002, 0.001) 167
[0.001, 0.01) 1049
[0.01, 0.1) 4553
[0.1, 1) 8840
[1, 10) 3822
[10, 100) 630
[100, 1e+03) 267
[1e+03, 1e+04) 699
[1e+04, 1e+05) 291
[1e+05, 1e+06) 83
[1e+06, 1e+07) 19
[1e+07, 2.33e+07] 19
-------------------------------------------------------------------------------
[0.1, 1) 386
[1, 10) 74
[10, 100) 101 456
[100, 1000) 34
[1000, 10000) 15
[100000, 1e+06] 1 1
0 974 323
[0.001, 0.01) 19
[0.1, 1) 370 57
[1, 10) 41 704
[10, 100] 2 246
-------------------------------------------------------------------------------
J.3 Problem with both linear and quadratic constraints
lower bd: 40 upper bd: 1 cont: all
upper bd: 121 fixed : 204
fixed : 5480 free : 5600
ranged : 161 ranged : 40
-------------------------------------------------------------------------------
Objective, maximize cx
range: all |c| in {0.00000, 15.4737}
distrib: |c| vars
0 5844
15.4737 1
-------------------------------------------------------------------------------
Row nonzeros, A_i
range: min A_i: 0 (0%) max A_i: 3 (0.0513259%)
0 80 1.38 1.38
1 5003 86.23 87.61
2 680 11.72 99.33
3 39 0.67 100.00
0/80 empty rows have quadratic terms
range: min A|j: 0 (0%) max A|j: 15 (0.258532%)
0 204 3.49 3.49
1 5521 94.46 97.95
2 40 0.68 98.63
[3, 7] 40 0.68 99.32
[8, 15] 40 0.68 100.00
0/204 empty columns correspond to variables used in conic
and/or quadratic expressions only
A nonzeros, A(ij)
range: min |A(ij)|: 2.02410e-05 max |A(ij)|: 35.8400
[2.02e-05, 0.0001) 40
[0.0001, 0.001) 118
[0.001, 0.01) 305
[0.01, 0.1) 176
[0.1, 1) 40
[1, 10) 5721
[10, 35.8] 80
-------------------------------------------------------------------------------
0 5481 5600
[1000, 10000) 1
[10000, 100000) 2 1
[1e+06, 1e+07) 78 40
[1e+08, 1e+09] 120 120
J.4. PROBLEM WITH BOTH LINEAR AND CONIC CONSTRAINTS 367
0 243 203
[0.1, 1) 1 1
[1e+06, 1e+07) 40
[1e+11, 1e+12] 1
-------------------------------------------------------------------------------
Quadratic constraints: 121
Gradient nonzeros, Qx
range: min Qx: 1 (0.0171086%) max Qx: 2720 (46.5355%)
distrib: Qx cons cons% acc%
1 40 33.06 33.06
[64, 127] 80 66.12 99.17
[2048, 2720] 1 0.83 100.00
-------------------------------------------------------------------------------
J.4 Problem with both linear and conic constraints
upper bd: 3600 fixed : 3601 cont: all
fixed : 21760 free : 28802
-------------------------------------------------------------------------------
Objective, minimize cx
range: all |c| in {0.00000, 1.00000}
distrib: |c| vars
0 32402
1 1
-------------------------------------------------------------------------------
Row nonzeros, A_i
1 3600 14.20 14.20
2 10803 42.60 56.79
[3, 7] 3995 15.75 72.55
8 6962 27.45 100.00
range: min A|j: 0 (0%) max A|j: 61 (0.240536%)
0 3602 11.12 11.12
1 10800 33.33 44.45
2 7200 22.22 66.67
[3, 7] 7279 22.46 89.13
[8, 15] 3521 10.87 100.00
[32, 61] 1 0.00 100.00
3600/3602 empty columns correspond to variables used in conic
and/or quadratic constraints only
A nonzeros, A(ij)
range: min |A(ij)|: 0.00833333 max |A(ij)|: 1.00000
[0.00833, 0.01) 57280
[0.01, 0.1) 59
[0.1, 1] 36000
-------------------------------------------------------------------------------
0 21760 21760
[0.1, 1] 3600
[1, 10] 3601 3601
-------------------------------------------------------------------------------
-------------------------------------------------------------------------------
Rotated quadratic cones: 3600
dim RQCs
4 3600
Bibliography
[1] Richard C. Grinold abd Ronald N. Kahn. Active portfolio management. McGraw-Hill, New York,
2 edition, 2000.
[2] F. Alizadeh and D. Goldfarb. Second-order cone programming. Math. Programming, 95(1):351,
2003.
[3] E. D. Andersen and K. D. Andersen. Presolving in linear programming. Math. Programming,
71(2):221245, 1995.
[4] E. D. Andersen, J. Gondzio, Cs. Meszaros, and X. Xu. Implementation of interior point methods
for large scale linear programming. In T. Terlaky, editor, Interior-point methods of mathematical
programming, pages 189252. Kluwer Academic Publishers, 1996.
[5] E. D. Andersen, C. Roos, and T. Terlaky. On implementing a primal-dual interior-point method
for conic quadratic optimization. Math. Programming, 95(2), February 2003.
[6] E. D. Andersen and Y. Ye. Combining interior-point and pivoting algorithms. Management Sci.,
42(12):17191731, December 1996.
[7] E. D. Andersen and Y. Ye. A computational study of the homogeneous algorithm for large-scale
convex optimization. Computational Optimization and Applications, 10:243269, 1998.
[8] E. D. Andersen and Y. Ye. On a homogeneous algorithm for the monotone complementarity
problem. Math. Programming, 84(2):375399, February 1999.
[9] Erling D. Andersen. The homogeneous and self-dual model and algorithm for linear optimization.
Technical Report TR-1-2009, MOSEK ApS, 2009. http://www.mosek.com/fileadmin/reports/
tech/homolo.pdf.
[10] M. S. Bazaraa, H. D. Sherali, and C. M. Shetty. Nonlinear programming: Theory and algorithms.
John Wiley and Sons, New York, 2 edition, 1993.
[11] A. Ben-Tal and A Nemirovski. Lectures on Modern Convex Optimization: Analysis, Algorithms,
and Engineering Applications. MPS/SIAM Series on Optimization. SIAM, 2001.
[12] V. Chvatal. Linear programming. W.H. Freeman and Company, 1983.
[13] R. Fourer, D. M. Gay, and B. W. Kernighan. AMPL. A modeling language for mathematical
programming. Thomson, 2nd edition, 2003.
369
370 BIBLIOGRAPHY
[14] N. Gould and P. L. Toint. Preprocessing for quadratic programming. Math. Programming,
100(1):95132, 2004.
[15] J. L. Kenningon and K. R. Lewis. Generalized networks: The theory of preprocessing and an
emperical analysis. INFORMS Journal on Computing, 16(2):162173, 2004.
[16] M. S. Lobo, L. Vanderberghe, S. Boyd, and H. Lebret. Applications of second-order cone pro-
gramming. Linear Algebra Appl., 284:193228, November 1998.
[17] M. S. Lobo and M. Fazel, and S. Boyd. Portfolio optimization with linear and xed transaction
costs. Technical report, CDS, California Institute of Technology, 2005. To appear in Annals of
Operations Research. http://www.cds.caltech.edu/~maryam/portfolio.html.
[18] J. L. Nazareth. Computer Solution of Linear Programs. Oxford University Press, New York, 1987.
[19] C. Roos, T. Terlaky, and J. -Ph. Vial. Theory and algorithms for linear optimization: an interior
point approach. John Wiley and Sons, New York, 1997.
[20] Bernd Scherer. Portfolio construction and risk budgeting. Risk Books, 2 edition, 2004.
[21] S. W. Wallace. Decision making under uncertainty: Is sensitivity of any use. Oper. Res., 48(1):20
25, January 2000.
[22] H. P. Williams. Model building in mathematical programming. John Wiley and Sons, 3 edition,
1993.
[23] L. A. Wolsey. Integer programming. John Wiley and Sons, 1998.
Index
absolute value, 50
alloc add qnz (parameter), 209
AMPL
outlev, 23
wantsol, 23
ana sol basis (parameter), 209
ana sol infeas tol (parameter), 178
ana sol print violated (parameter), 209
arguments
command line tool, 105
auto sort a before opt (parameter), 210
auto update sol info (parameter), 210
bas sol file name (parameter), 273
basis identication, 61
basis rel tol s (parameter), 178
basis solve use plus one (parameter), 210
basis tol s (parameter), 179
basis tol x (parameter), 179
bi clean optimizer (parameter), 211
bi ignore max iter (parameter), 211
bi ignore num error (parameter), 211
bi max iterations (parameter), 212
bounds, innite, 36
cache license (parameter), 212
cache size l1 (parameter), 212
cache size l2 (parameter), 213
callback freq (parameter), 179
certicate
dual, 38
primal, 37
check convexity (parameter), 213
check convexity rel tol (parameter), 179
check task data (parameter), 213
command line tool, 11, 105
complementarity conditions, 37
concurrent optimization, 67
concurrent solution, 67
concurrent num optimizers (parameter), 214
concurrent priority dual simplex (parameter), 214
concurrent priority free simplex (parameter), 214
concurrent priority intpnt (parameter), 214
concurrent priority primal simplex (parameter), 214
conic
optimization, 40
problem, 40
conic modelling, 42
minimizing norms, example, 43
pitfalls, 46
quadratic objective, example, 42
risk and market impact, example
Markowitz model, example, 51
constraint
matrix, 35, 48, 109
quadratic, 38, 39
constraints
lower limit, 35, 48, 109
upper limit, 35, 48, 109
continuous relaxation, 73
cpu type (parameter), 215
data check (parameter), 215
data file name (parameter), 273
data tol aij (parameter), 180
data tol aij huge (parameter), 180
data tol aij large (parameter), 180
data tol bound inf (parameter), 181
data tol bound wrn (parameter), 181
data tol c huge (parameter), 181
data tol cj large (parameter), 181
data tol qij (parameter), 182
data tol x (parameter), 182
debug file name (parameter), 274
dual certicate, 38
dual infeasible, 36, 38
duality gap (linear problem), 37
dualizer, 57
eliminator, 56
Embedded network ow problems, 64
371
372 INDEX
feasible, primal, 36
feasrepair name prefix (parameter), 274
feasrepair name separator (parameter), 274
feasrepair name wsumviol (parameter), 274
feasrepair optimize (parameter), 216
feasrepair tol (parameter), 182
help desk, 9
hot-start, 62
infeas generic names (parameter), 216
infeas prefer primal (parameter), 216
infeas report auto (parameter), 216
infeas report level (parameter), 217
infeasible, 83
dual, 38
primal, 37
infeasible problems, 83
infeasible, dual, 36
infeasible, primal, 36
innite bounds, 36
int sol file name (parameter), 275
integer optimization, 73
relaxation, 73
interior-point optimizer, 58, 65
interior-point or simplex optimizer, 63
intpnt basis (parameter), 217
intpnt co tol dfeas (parameter), 182
intpnt co tol infeas (parameter), 183
intpnt co tol mu red (parameter), 183
intpnt co tol near rel (parameter), 183
intpnt co tol pfeas (parameter), 184
intpnt co tol rel gap (parameter), 184
intpnt diff step (parameter), 218
intpnt factor debug lvl (parameter), 218
intpnt factor method (parameter), 218
intpnt max iterations (parameter), 218
intpnt max num cor (parameter), 219
intpnt max num refinement steps (parameter), 219
intpnt nl merit bal (parameter), 184
intpnt nl tol dfeas (parameter), 184
intpnt nl tol mu red (parameter), 185
intpnt nl tol near rel (parameter), 185
intpnt nl tol pfeas (parameter), 185
intpnt nl tol rel gap (parameter), 185
intpnt nl tol rel step (parameter), 186
intpnt num threads (parameter), 219
intpnt off col trh (parameter), 219
intpnt order method (parameter), 220
intpnt regularization use (parameter), 220
intpnt scaling (parameter), 220
intpnt solve form (parameter), 221
intpnt starting point (parameter), 221
intpnt tol dfeas (parameter), 186
intpnt tol dsafe (parameter), 186
intpnt tol infeas (parameter), 186
intpnt tol mu red (parameter), 187
intpnt tol path (parameter), 187
intpnt tol pfeas (parameter), 187
intpnt tol psafe (parameter), 187
intpnt tol rel gap (parameter), 188
intpnt tol rel step (parameter), 188
intpnt tol step size (parameter), 188
itr sol file name (parameter), 275
lic trh expiry wrn (parameter), 221
license allow overuse (parameter), 222
license cache time (parameter), 222
license check time (parameter), 222
license debug (parameter), 222
license pause time (parameter), 223
license suppress expire wrns (parameter), 223
license wait (parameter), 223
linear dependency check, 56
linear problem, 35
linearity interval, 96
log (parameter), 224
log bi (parameter), 224
log bi freq (parameter), 224
log check convexity (parameter), 224
log concurrent (parameter), 225
log cut second opt (parameter), 225
log factor (parameter), 225
log feasrepair (parameter), 226
log file (parameter), 226
log head (parameter), 226
log infeas ana (parameter), 226
log intpnt (parameter), 227
log mio (parameter), 227
log mio freq (parameter), 227
log nonconvex (parameter), 227
log optimizer (parameter), 228
log order (parameter), 228
log param (parameter), 228
log presolve (parameter), 228
log response (parameter), 229
log sensitivity (parameter), 229
log sensitivity opt (parameter), 229
log sim (parameter), 229
log sim freq (parameter), 230
log sim minor (parameter), 230
INDEX 373
log sim network freq (parameter), 230
log storage (parameter), 231
lower obj cut (parameter), 189
lower obj cut finite trh (parameter), 189
LP format, 121
lp write ignore incompatible items (parameter), 231
max num warnings (parameter), 231
mio branch dir (parameter), 231
mio branch priorities use (parameter), 232
mio construct sol (parameter), 232
mio cont sol (parameter), 232
mio cut level root (parameter), 233
mio cut level tree (parameter), 233
mio disable term time (parameter), 189
mio feaspump level (parameter), 233
mio heuristic level (parameter), 234
mio heuristic time (parameter), 190
mio hotstart (parameter), 234
mio keep basis (parameter), 234
mio local branch number (parameter), 235
mio max num branches (parameter), 235
mio max num relaxs (parameter), 235
mio max num solutions (parameter), 236
mio max time (parameter), 190
mio max time aprx opt (parameter), 190
mio mode (parameter), 236
mio near tol abs gap (parameter), 191
mio near tol rel gap (parameter), 191
mio node optimizer (parameter), 236
mio node selection (parameter), 237
mio optimizer mode (parameter), 237
mio presolve aggregate (parameter), 238
mio presolve probing (parameter), 238
mio presolve use (parameter), 238
mio rel add cut limited (parameter), 191
mio rel gap const (parameter), 191
mio root optimizer (parameter), 238
mio strong branch (parameter), 239
mio tol abs gap (parameter), 192
mio tol abs relax int (parameter), 192
mio tol feas (parameter), 192
mio tol rel gap (parameter), 193
mio tol rel relax int (parameter), 193
mio tol x (parameter), 193
mixed-integer optimization, 73
modelling
absolute value, 50
in cones, 42
market impact term, 52
Markowitz portfolio optimization, 52
minimizing a sum of norms, 43
portfolio optimization, 51
transaction costs, 52
MPS format, 109
BOUNDS, 115
COLUMNS, 112
free, 119
NAME, 111
OBJNAME, 111
OBJSENSE, 111
QSECTION, 114
RANGES, 113
RHS, 113
ROWS, 112
Network ow problems
embedded, 64
optimizing, 64
nonconvex max iterations (parameter), 239
nonconvex tol feas (parameter), 193
nonconvex tol opt (parameter), 193
objective
quadratic, 38
vector, 35
objective vector, 48
objective sense (parameter), 239
OPF format, 129
opf max terms per line (parameter), 240
opf write header (parameter), 240
opf write hints (parameter), 240
opf write parameters (parameter), 240
opf write problem (parameter), 241
opf write sol bas (parameter), 241
opf write sol itg (parameter), 241
opf write sol itr (parameter), 242
opf write solutions (parameter), 242
optimal solution, 37
optimization
conic, 40
integer, 73
mixed-integer, 73
optimizer (parameter), 242
optimizer max time (parameter), 194
optimizers
concurrent, 67
conic interior-point, 65
convex interior-point, 65
linear interior-point, 58
parallel, 67
374 INDEX
simplex, 62
Optimizing
network ow problems, 64
ORD format, 145
parallel extensions, 67
parallel interior-point, 57
parallel optimizers
interior point, 57
parallel solution, 67
param comment sign (parameter), 275
param read case name (parameter), 243
param read file name (parameter), 275
param read ign error (parameter), 243
param write file name (parameter), 276
parameter le, 107
parameters, 147
presolve, 55
eliminator, 56
linear dependency check, 56
presolve elim fill (parameter), 243
presolve eliminator max num tries (parameter), 244
presolve eliminator use (parameter), 244
presolve level (parameter), 244
presolve lindep use (parameter), 244
presolve lindep work lim (parameter), 245
presolve tol aij (parameter), 194
presolve tol lin dep (parameter), 194
presolve tol s (parameter), 194
presolve tol x (parameter), 195
presolve use (parameter), 245
primal feasible, 36
primal certicate, 37
primal infeasible, 36, 37
primal-dual solution, 36
qcqo reformulate rel drop tol (parameter), 195
qo separable reformulation (parameter), 245
quadratic constraint, 38, 39
quadratic objective, 38
quadratic optimization, 38
read add anz (parameter), 245
read add con (parameter), 246
read add cone (parameter), 246
read add qnz (parameter), 246
read add var (parameter), 246
read anz (parameter), 247
read con (parameter), 247
read cone (parameter), 247
read data compressed (parameter), 247
read data format (parameter), 248
read keep free con (parameter), 248
read lp drop new vars in bou (parameter), 248
read lp quoted names (parameter), 249
read mps bou name (parameter), 276
read mps format (parameter), 249
read mps keep int (parameter), 249
read mps obj name (parameter), 276
read mps obj sense (parameter), 250
read mps quoted names (parameter), 250
read mps ran name (parameter), 276
read mps relax (parameter), 250
read mps rhs name (parameter), 277
read mps width (parameter), 251
read q mode (parameter), 251
read qnz (parameter), 251
read task ignore param (parameter), 251
read var (parameter), 252
relaxation, continuous, 73
scaling, 57
sensitivity analysis, 95
basis type, 97
optimal partition type, 97
sensitivity all (parameter), 252
sensitivity file name (parameter), 277
sensitivity optimizer (parameter), 252
sensitivity res file name (parameter), 277
sensitivity type (parameter), 253
shadow price, 96
sim basis factor use (parameter), 253
sim degen (parameter), 253
sim dual crash (parameter), 254
sim dual phaseone method (parameter), 254
sim dual restrict selection (parameter), 254
sim dual selection (parameter), 255
sim exploit dupvec (parameter), 255
sim hotstart (parameter), 255
sim hotstart lu (parameter), 256
sim integer (parameter), 256
sim lu tol rel piv (parameter), 195
sim max iterations (parameter), 256
sim max num setbacks (parameter), 257
sim network detect (parameter), 257
sim network detect hotstart (parameter), 257
sim network detect method (parameter), 258
sim non singular (parameter), 258
sim primal crash (parameter), 258
sim primal phaseone method (parameter), 258
sim primal restrict selection (parameter), 259
INDEX 375
sim primal selection (parameter), 259
sim refactor freq (parameter), 260
sim reformulation (parameter), 260
sim save lu (parameter), 260
sim scaling (parameter), 261
sim scaling method (parameter), 261
sim solve form (parameter), 261
sim stability priority (parameter), 261
sim switch optimizer (parameter), 262
simplex optimizer, 62
simplex abs tol piv (parameter), 196
sol filter keep basic (parameter), 262
sol filter keep ranged (parameter), 262
sol filter xc low (parameter), 277
sol filter xc upr (parameter), 278
sol filter xx low (parameter), 278
sol filter xx upr (parameter), 278
sol quoted names (parameter), 263
sol read name width (parameter), 263
sol read width (parameter), 263
solution, optimal, 37
solution, primal-dual, 36
solution callback (parameter), 263
stat file name (parameter), 279
stat key (parameter), 279
stat name (parameter), 279
symbolic constants
MSK ACC CON, 281
MSK ACC VAR, 281
MSK ADOP ADD, 281
MSK ADOP DIV, 281
MSK ADOP EXP, 281
MSK ADOP LOG, 281
MSK ADOP MUL, 281
MSK ADOP POW, 281
MSK ADOP RET, 281
MSK ADOP SUB, 281
MSK ADOPTYPE CONSTANT, 282
MSK ADOPTYPE NONE, 282
MSK ADOPTYPE REFERENCE, 282
MSK ADOPTYPE VARIABLE, 282
MSK BI ALWAYS, 282
MSK BI IF FEASIBLE, 282
MSK BI NEVER, 282
MSK BI NO ERROR, 282
MSK BI OTHER, 282
MSK BK FR, 283
MSK BK FX, 282
MSK BK LO, 283
MSK BK RA, 283
MSK BK UP, 283
MSK BRANCH DIR DOWN, 283
MSK BRANCH DIR FREE, 283
MSK BRANCH DIR UP, 283
MSK CALLBACK BEGIN BI, 290
MSK CALLBACK BEGIN CONCURRENT, 285
MSK CALLBACK BEGIN CONIC, 285
MSK CALLBACK BEGIN DUAL BI, 284
MSK CALLBACK BEGIN DUAL SENSITIVITY, 284
MSK CALLBACK BEGIN DUAL SETUP BI, 286
MSK CALLBACK BEGIN DUAL SIMPLEX, 288
MSK CALLBACK BEGIN DUAL SIMPLEX BI, 289
MSK CALLBACK BEGIN FULL CONVEXITY CHECK, 284
MSK CALLBACK BEGIN INFEAS ANA, 289
MSK CALLBACK BEGIN INTPNT, 289
MSK CALLBACK BEGIN LICENSE WAIT, 285
MSK CALLBACK BEGIN MIO, 290
MSK CALLBACK BEGIN NETWORK DUAL SIMPLEX, 284
MSK CALLBACK BEGIN NETWORK PRIMAL SIMPLEX, 291
MSK CALLBACK BEGIN NETWORK SIMPLEX, 285
MSK CALLBACK BEGIN NONCONVEX, 287
MSK CALLBACK BEGIN OPTIMIZER, 288
MSK CALLBACK BEGIN PRESOLVE, 290
MSK CALLBACK BEGIN PRIMAL BI, 284
MSK CALLBACK BEGIN PRIMAL DUAL SIMPLEX, 289
MSK CALLBACK BEGIN PRIMAL DUAL SIMPLEX BI, 283
MSK CALLBACK BEGIN PRIMAL SENSITIVITY, 289
MSK CALLBACK BEGIN PRIMAL SETUP BI, 288
MSK CALLBACK BEGIN PRIMAL SIMPLEX, 288
MSK CALLBACK BEGIN PRIMAL SIMPLEX BI, 287
MSK CALLBACK BEGIN QCQO REFORMULATE, 286
MSK CALLBACK BEGIN READ, 288
MSK CALLBACK BEGIN SIMPLEX, 289
MSK CALLBACK BEGIN SIMPLEX BI, 289
MSK CALLBACK BEGIN SIMPLEX NETWORK DETECT, 287
MSK CALLBACK BEGIN WRITE, 288
MSK CALLBACK CONIC, 285
MSK CALLBACK DUAL SIMPLEX, 287
MSK CALLBACK END BI, 288
MSK CALLBACK END CONCURRENT, 284
MSK CALLBACK END CONIC, 286
MSK CALLBACK END DUAL BI, 287
MSK CALLBACK END DUAL SENSITIVITY, 286
MSK CALLBACK END DUAL SETUP BI, 288
MSK CALLBACK END DUAL SIMPLEX, 289
MSK CALLBACK END DUAL SIMPLEX BI, 286
MSK CALLBACK END FULL CONVEXITY CHECK, 290
MSK CALLBACK END INFEAS ANA, 286
MSK CALLBACK END INTPNT, 283
MSK CALLBACK END LICENSE WAIT, 286
376 INDEX
MSK CALLBACK END MIO, 284
MSK CALLBACK END NETWORK DUAL SIMPLEX, 284
MSK CALLBACK END NETWORK PRIMAL SIMPLEX, 283
MSK CALLBACK END NETWORK SIMPLEX, 285
MSK CALLBACK END NONCONVEX, 288
MSK CALLBACK END OPTIMIZER, 285
MSK CALLBACK END PRESOLVE, 289
MSK CALLBACK END PRIMAL BI, 290
MSK CALLBACK END PRIMAL DUAL SIMPLEX, 290
MSK CALLBACK END PRIMAL DUAL SIMPLEX BI, 288
MSK CALLBACK END PRIMAL SENSITIVITY, 289
MSK CALLBACK END PRIMAL SETUP BI, 288
MSK CALLBACK END PRIMAL SIMPLEX, 288
MSK CALLBACK END PRIMAL SIMPLEX BI, 286
MSK CALLBACK END QCQO REFORMULATE, 289
MSK CALLBACK END READ, 287
MSK CALLBACK END SIMPLEX, 290
MSK CALLBACK END SIMPLEX BI, 289
MSK CALLBACK END SIMPLEX NETWORK DETECT, 284
MSK CALLBACK END WRITE, 289
MSK CALLBACK IM BI, 288
MSK CALLBACK IM CONIC, 290
MSK CALLBACK IM DUAL BI, 284
MSK CALLBACK IM DUAL SENSIVITY, 289
MSK CALLBACK IM DUAL SIMPLEX, 285
MSK CALLBACK IM FULL CONVEXITY CHECK, 287
MSK CALLBACK IM INTPNT, 284
MSK CALLBACK IM LICENSE WAIT, 285
MSK CALLBACK IM LU, 286
MSK CALLBACK IM MIO, 290
MSK CALLBACK IM MIO DUAL SIMPLEX, 290
MSK CALLBACK IM MIO INTPNT, 290
MSK CALLBACK IM MIO PRESOLVE, 290
MSK CALLBACK IM MIO PRIMAL SIMPLEX, 284
MSK CALLBACK IM NETWORK DUAL SIMPLEX, 288
MSK CALLBACK IM NETWORK PRIMAL SIMPLEX, 285
MSK CALLBACK IM NONCONVEX, 288
MSK CALLBACK IM ORDER, 290
MSK CALLBACK IM PRESOLVE, 286
MSK CALLBACK IM PRIMAL BI, 290
MSK CALLBACK IM PRIMAL DUAL SIMPLEX, 287
MSK CALLBACK IM PRIMAL SENSIVITY, 289
MSK CALLBACK IM PRIMAL SIMPLEX, 284
MSK CALLBACK IM QO REFORMULATE, 285
MSK CALLBACK IM SIMPLEX, 285
MSK CALLBACK IM SIMPLEX BI, 287
MSK CALLBACK INTPNT, 286
MSK CALLBACK NEW INT MIO, 284
MSK CALLBACK NONCOVEX, 286
MSK CALLBACK PRIMAL SIMPLEX, 287
MSK CALLBACK QCONE, 289
MSK CALLBACK READ ADD ANZ, 287
MSK CALLBACK READ ADD CONES, 287
MSK CALLBACK READ ADD CONS, 283
MSK CALLBACK READ ADD QNZ, 285
MSK CALLBACK READ ADD VARS, 286
MSK CALLBACK READ OPF, 285
MSK CALLBACK READ OPF SECTION, 286
MSK CALLBACK UPDATE DUAL BI, 285
MSK CALLBACK UPDATE DUAL SIMPLEX, 290
MSK CALLBACK UPDATE DUAL SIMPLEX BI, 285
MSK CALLBACK UPDATE NETWORK DUAL SIMPLEX, 286
MSK CALLBACK UPDATE NETWORK PRIMAL SIMPLEX, 290
MSK CALLBACK UPDATE NONCONVEX, 287
MSK CALLBACK UPDATE PRESOLVE, 285
MSK CALLBACK UPDATE PRIMAL BI, 286
MSK CALLBACK UPDATE PRIMAL DUAL SIMPLEX, 289
MSK CALLBACK UPDATE PRIMAL DUAL SIMPLEX BI, 286
MSK CALLBACK UPDATE PRIMAL SIMPLEX, 287
MSK CALLBACK UPDATE PRIMAL SIMPLEX BI, 283
MSK CALLBACK WRITE OPF, 290
MSK CHECK CONVEXITY FULL, 291
MSK CHECK CONVEXITY NONE, 291
MSK CHECK CONVEXITY SIMPLE, 291
MSK COMPRESS FREE, 291
MSK COMPRESS GZIP, 291
MSK COMPRESS NONE, 291
MSK CPU AMD ATHLON, 292
MSK CPU AMD OPTERON, 292
MSK CPU GENERIC, 292
MSK CPU HP PARISC20, 292
MSK CPU INTEL CORE2, 292
MSK CPU INTEL ITANIUM2, 292
MSK CPU INTEL P3, 292
MSK CPU INTEL P4, 292
MSK CPU INTEL PM, 292
MSK CPU POWERPC G5, 292
MSK CPU UNKNOWN, 292
MSK CT QUAD, 291
MSK CT RQUAD, 291
MSK DATA FORMAT EXTENSION, 292
MSK DATA FORMAT FREE MPS, 292
MSK DATA FORMAT LP, 292
MSK DATA FORMAT MBT, 292
MSK DATA FORMAT MPS, 292
MSK DATA FORMAT OP, 292
MSK DATA FORMAT XML, 292
MSK DINF BI CLEAN DUAL TIME, 296
MSK DINF BI CLEAN PRIMAL DUAL TIME, 296
MSK DINF BI CLEAN PRIMAL TIME, 297
INDEX 377
MSK DINF BI CLEAN TIME, 295
MSK DINF BI DUAL TIME, 293
MSK DINF BI PRIMAL TIME, 296
MSK DINF BI TIME, 296
MSK DINF CONCURRENT TIME, 295
MSK DINF INTPNT DUAL FEAS, 295
MSK DINF INTPNT DUAL OBJ, 294
MSK DINF INTPNT FACTOR NUM FLOPS, 293
MSK DINF INTPNT KAP DIV TAU, 295
MSK DINF INTPNT ORDER TIME, 296
MSK DINF INTPNT PRIMAL FEAS, 293
MSK DINF INTPNT PRIMAL OBJ, 296
MSK DINF INTPNT TIME, 294
MSK DINF MIO CONSTRUCT SOLUTION OBJ, 294
MSK DINF MIO HEURISTIC TIME, 294
MSK DINF MIO OBJ ABS GAP, 296
MSK DINF MIO OBJ BOUND, 296
MSK DINF MIO OBJ INT, 295
MSK DINF MIO OBJ REL GAP, 293
MSK DINF MIO OPTIMIZER TIME, 293
MSK DINF MIO ROOT OPTIMIZER TIME, 297
MSK DINF MIO ROOT PRESOLVE TIME, 294
MSK DINF MIO TIME, 293
MSK DINF MIO USER OBJ CUT, 296
MSK DINF OPTIMIZER TIME, 294
MSK DINF PRESOLVE ELI TIME, 293
MSK DINF PRESOLVE LINDEP TIME, 293
MSK DINF PRESOLVE TIME, 293
MSK DINF QCQO REFORMULATE TIME, 295
MSK DINF RD TIME, 293
MSK DINF SIM DUAL TIME, 293
MSK DINF SIM FEAS, 294
MSK DINF SIM NETWORK DUAL TIME, 295
MSK DINF SIM NETWORK PRIMAL TIME, 295
MSK DINF SIM NETWORK TIME, 293
MSK DINF SIM OBJ, 293
MSK DINF SIM PRIMAL DUAL TIME, 297
MSK DINF SIM PRIMAL TIME, 294
MSK DINF SIM TIME, 294
MSK DINF SOL BAS DUAL OBJ, 296
MSK DINF SOL BAS MAX DBI, 296
MSK DINF SOL BAS MAX DEQI, 297
MSK DINF SOL BAS MAX PBI, 294
MSK DINF SOL BAS MAX PEQI, 296
MSK DINF SOL BAS MAX PINTI, 295
MSK DINF SOL BAS PRIMAL OBJ, 294
MSK DINF SOL INT MAX PBI, 295
MSK DINF SOL INT MAX PEQI, 295
MSK DINF SOL INT MAX PINTI, 295
MSK DINF SOL INT PRIMAL OBJ, 296
MSK DINF SOL ITR DUAL OBJ, 295
MSK DINF SOL ITR MAX DBI, 297
MSK DINF SOL ITR MAX DCNI, 294
MSK DINF SOL ITR MAX DEQI, 294
MSK DINF SOL ITR MAX PBI, 294
MSK DINF SOL ITR MAX PCNI, 293
MSK DINF SOL ITR MAX PEQI, 297
MSK DINF SOL ITR MAX PINTI, 293
MSK DINF SOL ITR PRIMAL OBJ, 295
MSK DPAR ANA SOL INFEAS TOL, 298
MSK DPAR BASIS REL TOL S, 301
MSK DPAR BASIS TOL S, 297
MSK DPAR BASIS TOL X, 302
MSK DPAR CALLBACK FREQ, 302
MSK DPAR CHECK CONVEXITY REL TOL, 298
MSK DPAR DATA TOL AIJ, 299
MSK DPAR DATA TOL AIJ HUGE, 302
MSK DPAR DATA TOL AIJ LARGE, 298
MSK DPAR DATA TOL BOUND INF, 302
MSK DPAR DATA TOL BOUND WRN, 302
MSK DPAR DATA TOL C HUGE, 301
MSK DPAR DATA TOL CJ LARGE, 301
MSK DPAR DATA TOL QIJ, 303
MSK DPAR DATA TOL X, 300
MSK DPAR FEASREPAIR TOL, 299
MSK DPAR INTPNT CO TOL DFEAS, 298
MSK DPAR INTPNT CO TOL INFEAS, 301
MSK DPAR INTPNT CO TOL MU RED, 300
MSK DPAR INTPNT CO TOL NEAR REL, 302
MSK DPAR INTPNT CO TOL PFEAS, 303
MSK DPAR INTPNT CO TOL REL GAP, 300
MSK DPAR INTPNT NL MERIT BAL, 302
MSK DPAR INTPNT NL TOL DFEAS, 302
MSK DPAR INTPNT NL TOL MU RED, 299
MSK DPAR INTPNT NL TOL NEAR REL, 299
MSK DPAR INTPNT NL TOL PFEAS, 302
MSK DPAR INTPNT NL TOL REL GAP, 303
MSK DPAR INTPNT NL TOL REL STEP, 301
MSK DPAR INTPNT TOL DFEAS, 302
MSK DPAR INTPNT TOL DSAFE, 299
MSK DPAR INTPNT TOL INFEAS, 299
MSK DPAR INTPNT TOL MU RED, 300
MSK DPAR INTPNT TOL PATH, 301
MSK DPAR INTPNT TOL PFEAS, 301
MSK DPAR INTPNT TOL PSAFE, 302
MSK DPAR INTPNT TOL REL GAP, 302
MSK DPAR INTPNT TOL REL STEP, 300
MSK DPAR INTPNT TOL STEP SIZE, 301
MSK DPAR LOWER OBJ CUT, 300
MSK DPAR LOWER OBJ CUT FINITE TRH, 297
378 INDEX
MSK DPAR MIO DISABLE TERM TIME, 300
MSK DPAR MIO HEURISTIC TIME, 298
MSK DPAR MIO MAX TIME, 297
MSK DPAR MIO MAX TIME APRX OPT, 301
MSK DPAR MIO NEAR TOL ABS GAP, 302
MSK DPAR MIO NEAR TOL REL GAP, 299
MSK DPAR MIO REL ADD CUT LIMITED, 300
MSK DPAR MIO REL GAP CONST, 301
MSK DPAR MIO TOL ABS GAP, 298
MSK DPAR MIO TOL ABS RELAX INT, 298
MSK DPAR MIO TOL FEAS, 300
MSK DPAR MIO TOL REL GAP, 302
MSK DPAR MIO TOL REL RELAX INT, 303
MSK DPAR MIO TOL X, 300
MSK DPAR NONCONVEX TOL FEAS, 298
MSK DPAR NONCONVEX TOL OPT, 298
MSK DPAR OPTIMIZER MAX TIME, 299
MSK DPAR PRESOLVE TOL AIJ, 301
MSK DPAR PRESOLVE TOL LIN DEP, 301
MSK DPAR PRESOLVE TOL S, 297
MSK DPAR PRESOLVE TOL X, 299
MSK DPAR QCQO REFORMULATE REL DROP TOL, 303
MSK DPAR SIM LU TOL REL PIV, 301
MSK DPAR SIMPLEX ABS TOL PIV, 299
MSK DPAR UPPER OBJ CUT, 297
MSK DPAR UPPER OBJ CUT FINITE TRH, 298
MSK FEASREPAIR OPTIMIZE COMBINED, 303
MSK FEASREPAIR OPTIMIZE NONE, 303
MSK FEASREPAIR OPTIMIZE PENALTY, 303
MSK FEATURE PTOM, 303
MSK FEATURE PTON, 303
MSK FEATURE PTOX, 303
MSK FEATURE PTS, 303
MSK IINF ANA PRO NUM CON, 304
MSK IINF ANA PRO NUM CON EQ, 307
MSK IINF ANA PRO NUM CON FR, 305
MSK IINF ANA PRO NUM CON LO, 307
MSK IINF ANA PRO NUM CON RA, 309
MSK IINF ANA PRO NUM CON UP, 305
MSK IINF ANA PRO NUM VAR, 306
MSK IINF ANA PRO NUM VAR BIN, 308
MSK IINF ANA PRO NUM VAR CONT, 304
MSK IINF ANA PRO NUM VAR EQ, 308
MSK IINF ANA PRO NUM VAR FR, 308
MSK IINF ANA PRO NUM VAR INT, 304
MSK IINF ANA PRO NUM VAR LO, 306
MSK IINF ANA PRO NUM VAR RA, 308
MSK IINF ANA PRO NUM VAR UP, 308
MSK IINF CACHE SIZE L1, 306
MSK IINF CACHE SIZE L2, 306
MSK IINF CONCURRENT FASTEST OPTIMIZER, 308
MSK IINF CPU TYPE, 307
MSK IINF INTPNT FACTOR NUM OFFCOL, 304
MSK IINF INTPNT ITER, 305
MSK IINF INTPNT NUM THREADS, 308
MSK IINF INTPNT SOLVE DUAL, 305
MSK IINF MIO CONSTRUCT SOLUTION, 306
MSK IINF MIO INITIAL SOLUTION, 308
MSK IINF MIO NUM ACTIVE NODES, 308
MSK IINF MIO NUM BRANCH, 307
MSK IINF MIO NUM CUTS, 306
MSK IINF MIO NUM INT SOLUTIONS, 307
MSK IINF MIO NUM RELAX, 307
MSK IINF MIO NUMCON, 304
MSK IINF MIO NUMINT, 306
MSK IINF MIO NUMVAR, 306
MSK IINF MIO TOTAL NUM BASIS CUTS, 305
MSK IINF MIO TOTAL NUM BRANCH, 307
MSK IINF MIO TOTAL NUM CARDGUB CUTS, 306
MSK IINF MIO TOTAL NUM CLIQUE CUTS, 305
MSK IINF MIO TOTAL NUM COEF REDC CUTS, 309
MSK IINF MIO TOTAL NUM CONTRA CUTS, 306
MSK IINF MIO TOTAL NUM CUTS, 309
MSK IINF MIO TOTAL NUM DISAGG CUTS, 309
MSK IINF MIO TOTAL NUM FLOW COVER CUTS, 307
MSK IINF MIO TOTAL NUM GCD CUTS, 307
MSK IINF MIO TOTAL NUM GOMORY CUTS, 309
MSK IINF MIO TOTAL NUM GUB COVER CUTS, 307
MSK IINF MIO TOTAL NUM KNAPSUR COVER CUTS, 305
MSK IINF MIO TOTAL NUM LATTICE CUTS, 304
MSK IINF MIO TOTAL NUM LIFT CUTS, 307
MSK IINF MIO TOTAL NUM OBJ CUTS, 304
MSK IINF MIO TOTAL NUM PLAN LOC CUTS, 306
MSK IINF MIO TOTAL NUM RELAX, 308
MSK IINF MIO USER OBJ CUT, 308
MSK IINF OPT NUMCON, 309
MSK IINF OPT NUMVAR, 304
MSK IINF OPTIMIZE RESPONSE, 305
MSK IINF RD NUMCON, 308
MSK IINF RD NUMCONE, 308
MSK IINF RD NUMINTVAR, 304
MSK IINF RD NUMQ, 304
MSK IINF RD NUMVAR, 306
MSK IINF RD PROTYPE, 305
MSK IINF SIM DUAL DEG ITER, 305
MSK IINF SIM DUAL HOTSTART, 308
MSK IINF SIM DUAL HOTSTART LU, 308
MSK IINF SIM DUAL INF ITER, 306
MSK IINF SIM DUAL ITER, 305
MSK IINF SIM NETWORK DUAL DEG ITER, 306
INDEX 379
MSK IINF SIM NETWORK DUAL HOTSTART, 305
MSK IINF SIM NETWORK DUAL HOTSTART LU, 307
MSK IINF SIM NETWORK DUAL INF ITER, 304
MSK IINF SIM NETWORK DUAL ITER, 308
MSK IINF SIM NETWORK PRIMAL DEG ITER, 304
MSK IINF SIM NETWORK PRIMAL HOTSTART, 309
MSK IINF SIM NETWORK PRIMAL HOTSTART LU, 308
MSK IINF SIM NETWORK PRIMAL INF ITER, 307
MSK IINF SIM NETWORK PRIMAL ITER, 307
MSK IINF SIM NUMCON, 305
MSK IINF SIM NUMVAR, 304
MSK IINF SIM PRIMAL DEG ITER, 309
MSK IINF SIM PRIMAL DUAL DEG ITER, 306
MSK IINF SIM PRIMAL DUAL HOTSTART, 305
MSK IINF SIM PRIMAL DUAL HOTSTART LU, 305
MSK IINF SIM PRIMAL DUAL INF ITER, 309
MSK IINF SIM PRIMAL DUAL ITER, 305
MSK IINF SIM PRIMAL HOTSTART, 307
MSK IINF SIM PRIMAL HOTSTART LU, 307
MSK IINF SIM PRIMAL INF ITER, 308
MSK IINF SIM PRIMAL ITER, 308
MSK IINF SIM SOLVE DUAL, 304
MSK IINF SOL BAS PROSTA, 307
MSK IINF SOL BAS SOLSTA, 304
MSK IINF SOL INT PROSTA, 306
MSK IINF SOL INT SOLSTA, 306
MSK IINF SOL ITR PROSTA, 305
MSK IINF SOL ITR SOLSTA, 305
MSK IINF STO NUM A CACHE FLUSHES, 306
MSK IINF STO NUM A REALLOC, 307
MSK IINF STO NUM A TRANSPOSES, 304
MSK INF DOU TYPE, 309
MSK INF INT TYPE, 309
MSK INF LINT TYPE, 309
MSK IOMODE READ, 309
MSK IOMODE READWRITE, 310
MSK IOMODE WRITE, 309
MSK IPAR ALLOC ADD QNZ, 322
MSK IPAR ANA SOL BASIS, 316
MSK IPAR ANA SOL PRINT VIOLATED, 315
MSK IPAR AUTO SORT A BEFORE OPT, 323
MSK IPAR AUTO UPDATE SOL INFO, 314
MSK IPAR BASIS SOLVE USE PLUS ONE, 314
MSK IPAR BI CLEAN OPTIMIZER, 320
MSK IPAR BI IGNORE MAX ITER, 319
MSK IPAR BI IGNORE NUM ERROR, 325
MSK IPAR BI MAX ITERATIONS, 317
MSK IPAR CACHE LICENSE, 319
MSK IPAR CACHE SIZE L1, 324
MSK IPAR CACHE SIZE L2, 324
MSK IPAR CHECK CONVEXITY, 316
MSK IPAR CHECK TASK DATA, 312
MSK IPAR CONCURRENT NUM OPTIMIZERS, 316
MSK IPAR CONCURRENT PRIORITY DUAL SIMPLEX, 313
MSK IPAR CONCURRENT PRIORITY FREE SIMPLEX, 321
MSK IPAR CONCURRENT PRIORITY INTPNT, 321
MSK IPAR CONCURRENT PRIORITY PRIMAL SIMPLEX,
318
MSK IPAR CPU TYPE, 315
MSK IPAR DATA CHECK, 321
MSK IPAR FEASREPAIR OPTIMIZE, 314
MSK IPAR INFEAS GENERIC NAMES, 324
MSK IPAR INFEAS PREFER PRIMAL, 320
MSK IPAR INFEAS REPORT AUTO, 310
MSK IPAR INFEAS REPORT LEVEL, 324
MSK IPAR INTPNT BASIS, 320
MSK IPAR INTPNT DIFF STEP, 319
MSK IPAR INTPNT FACTOR DEBUG LVL, 317
MSK IPAR INTPNT FACTOR METHOD, 325
MSK IPAR INTPNT MAX ITERATIONS, 315
MSK IPAR INTPNT MAX NUM COR, 315
MSK IPAR INTPNT MAX NUM REFINEMENT STEPS, 315
MSK IPAR INTPNT NUM THREADS, 322
MSK IPAR INTPNT OFF COL TRH, 313
MSK IPAR INTPNT ORDER METHOD, 311
MSK IPAR INTPNT REGULARIZATION USE, 320
MSK IPAR INTPNT SCALING, 326
MSK IPAR INTPNT SOLVE FORM, 319
MSK IPAR INTPNT STARTING POINT, 318
MSK IPAR LIC TRH EXPIRY WRN, 313
MSK IPAR LICENSE ALLOW OVERUSE, 320
MSK IPAR LICENSE CACHE TIME, 317
MSK IPAR LICENSE CHECK TIME, 321
MSK IPAR LICENSE DEBUG, 316
MSK IPAR LICENSE PAUSE TIME, 317
MSK IPAR LICENSE SUPPRESS EXPIRE WRNS, 323
MSK IPAR LICENSE WAIT, 316
MSK IPAR LOG, 318
MSK IPAR LOG BI, 315
MSK IPAR LOG BI FREQ, 313
MSK IPAR LOG CHECK CONVEXITY, 325
MSK IPAR LOG CONCURRENT, 315
MSK IPAR LOG CUT SECOND OPT, 320
MSK IPAR LOG FACTOR, 314
MSK IPAR LOG FEASREPAIR, 319
MSK IPAR LOG FILE, 321
MSK IPAR LOG HEAD, 325
MSK IPAR LOG INFEAS ANA, 311
MSK IPAR LOG INTPNT, 317
MSK IPAR LOG MIO, 317
380 INDEX
MSK IPAR LOG MIO FREQ, 312
MSK IPAR LOG NONCONVEX, 314
MSK IPAR LOG OPTIMIZER, 322
MSK IPAR LOG ORDER, 314
MSK IPAR LOG PARAM, 317
MSK IPAR LOG PRESOLVE, 313
MSK IPAR LOG RESPONSE, 324
MSK IPAR LOG SENSITIVITY, 311
MSK IPAR LOG SENSITIVITY OPT, 310
MSK IPAR LOG SIM, 313
MSK IPAR LOG SIM FREQ, 313
MSK IPAR LOG SIM MINOR, 315
MSK IPAR LOG SIM NETWORK FREQ, 319
MSK IPAR LOG STORAGE, 320
MSK IPAR LP WRITE IGNORE INCOMPATIBLE ITEMS,
316
MSK IPAR MAX NUM WARNINGS, 323
MSK IPAR MIO BRANCH DIR, 317
MSK IPAR MIO BRANCH PRIORITIES USE, 310
MSK IPAR MIO CONSTRUCT SOL, 322
MSK IPAR MIO CONT SOL, 318
MSK IPAR MIO CUT LEVEL ROOT, 314
MSK IPAR MIO CUT LEVEL TREE, 311
MSK IPAR MIO FEASPUMP LEVEL, 323
MSK IPAR MIO HEURISTIC LEVEL, 325
MSK IPAR MIO HOTSTART, 322
MSK IPAR MIO KEEP BASIS, 326
MSK IPAR MIO LOCAL BRANCH NUMBER, 312
MSK IPAR MIO MAX NUM BRANCHES, 322
MSK IPAR MIO MAX NUM RELAXS, 325
MSK IPAR MIO MAX NUM SOLUTIONS, 321
MSK IPAR MIO MODE, 313
MSK IPAR MIO NODE OPTIMIZER, 310
MSK IPAR MIO NODE SELECTION, 315
MSK IPAR MIO OPTIMIZER MODE, 317
MSK IPAR MIO PRESOLVE AGGREGATE, 317
MSK IPAR MIO PRESOLVE PROBING, 320
MSK IPAR MIO PRESOLVE USE, 320
MSK IPAR MIO ROOT OPTIMIZER, 312
MSK IPAR MIO STRONG BRANCH, 323
MSK IPAR NONCONVEX MAX ITERATIONS, 312
MSK IPAR OBJECTIVE SENSE, 318
MSK IPAR OPF MAX TERMS PER LINE, 319
MSK IPAR OPF WRITE HEADER, 318
MSK IPAR OPF WRITE HINTS, 322
MSK IPAR OPF WRITE PARAMETERS, 311
MSK IPAR OPF WRITE PROBLEM, 325
MSK IPAR OPF WRITE SOL BAS, 312
MSK IPAR OPF WRITE SOL ITG, 310
MSK IPAR OPF WRITE SOL ITR, 311
MSK IPAR OPF WRITE SOLUTIONS, 318
MSK IPAR OPTIMIZER, 323
MSK IPAR PARAM READ CASE NAME, 322
MSK IPAR PARAM READ IGN ERROR, 318
MSK IPAR PRESOLVE ELIM FILL, 312
MSK IPAR PRESOLVE ELIMINATOR MAX NUM TRIES, 316
MSK IPAR PRESOLVE ELIMINATOR USE, 322
MSK IPAR PRESOLVE LEVEL, 310
MSK IPAR PRESOLVE LINDEP USE, 318
MSK IPAR PRESOLVE LINDEP WORK LIM, 315
MSK IPAR PRESOLVE USE, 310
MSK IPAR QO SEPARABLE REFORMULATION, 316
MSK IPAR READ ADD ANZ, 311
MSK IPAR READ ADD CON, 315
MSK IPAR READ ADD CONE, 310
MSK IPAR READ ADD QNZ, 313
MSK IPAR READ ADD VAR, 310
MSK IPAR READ ANZ, 313
MSK IPAR READ CON, 311
MSK IPAR READ CONE, 323
MSK IPAR READ DATA COMPRESSED, 324
MSK IPAR READ DATA FORMAT, 325
MSK IPAR READ KEEP FREE CON, 321
MSK IPAR READ LP DROP NEW VARS IN BOU, 313
MSK IPAR READ LP QUOTED NAMES, 320
MSK IPAR READ MPS FORMAT, 322
MSK IPAR READ MPS KEEP INT, 321
MSK IPAR READ MPS OBJ SENSE, 319
MSK IPAR READ MPS QUOTED NAMES, 322
MSK IPAR READ MPS RELAX, 311
MSK IPAR READ MPS WIDTH, 314
MSK IPAR READ Q MODE, 315
MSK IPAR READ QNZ, 311
MSK IPAR READ TASK IGNORE PARAM, 316
MSK IPAR READ VAR, 325
MSK IPAR SENSITIVITY ALL, 319
MSK IPAR SENSITIVITY OPTIMIZER, 321
MSK IPAR SENSITIVITY TYPE, 312
MSK IPAR SIM BASIS FACTOR USE, 320
MSK IPAR SIM DEGEN, 322
MSK IPAR SIM DUAL CRASH, 325
MSK IPAR SIM DUAL PHASEONE METHOD, 323
MSK IPAR SIM DUAL RESTRICT SELECTION, 312
MSK IPAR SIM DUAL SELECTION, 316
MSK IPAR SIM EXPLOIT DUPVEC, 317
MSK IPAR SIM HOTSTART, 317
MSK IPAR SIM HOTSTART LU, 318
MSK IPAR SIM INTEGER, 323
MSK IPAR SIM MAX ITERATIONS, 315
MSK IPAR SIM MAX NUM SETBACKS, 321
INDEX 381
MSK IPAR SIM NETWORK DETECT, 324
MSK IPAR SIM NETWORK DETECT HOTSTART, 318
MSK IPAR SIM NETWORK DETECT METHOD, 325
MSK IPAR SIM NON SINGULAR, 321
MSK IPAR SIM PRIMAL CRASH, 322
MSK IPAR SIM PRIMAL PHASEONE METHOD, 310
MSK IPAR SIM PRIMAL RESTRICT SELECTION, 323
MSK IPAR SIM PRIMAL SELECTION, 311
MSK IPAR SIM REFACTOR FREQ, 316
MSK IPAR SIM REFORMULATION, 325
MSK IPAR SIM SAVE LU, 325
MSK IPAR SIM SCALING, 319
MSK IPAR SIM SCALING METHOD, 313
MSK IPAR SIM SOLVE FORM, 316
MSK IPAR SIM STABILITY PRIORITY, 310
MSK IPAR SIM SWITCH OPTIMIZER, 324
MSK IPAR SOL FILTER KEEP BASIC, 325
MSK IPAR SOL FILTER KEEP RANGED, 319
MSK IPAR SOL QUOTED NAMES, 317
MSK IPAR SOL READ NAME WIDTH, 316
MSK IPAR SOL READ WIDTH, 315
MSK IPAR SOLUTION CALLBACK, 313
MSK IPAR TIMING LEVEL, 314
MSK IPAR WARNING LEVEL, 312
MSK IPAR WRITE BAS CONSTRAINTS, 323
MSK IPAR WRITE BAS HEAD, 310
MSK IPAR WRITE BAS VARIABLES, 320
MSK IPAR WRITE DATA COMPRESSED, 311
MSK IPAR WRITE DATA FORMAT, 317
MSK IPAR WRITE DATA PARAM, 321
MSK IPAR WRITE FREE CON, 312
MSK IPAR WRITE GENERIC NAMES, 312
MSK IPAR WRITE GENERIC NAMES IO, 316
MSK IPAR WRITE INT CONSTRAINTS, 311
MSK IPAR WRITE INT HEAD, 318
MSK IPAR WRITE INT VARIABLES, 311
MSK IPAR WRITE LP LINE WIDTH, 316
MSK IPAR WRITE LP QUOTED NAMES, 314
MSK IPAR WRITE LP STRICT FORMAT, 312
MSK IPAR WRITE LP TERMS PER LINE, 319
MSK IPAR WRITE MPS INT, 321
MSK IPAR WRITE MPS OBJ SENSE, 322
MSK IPAR WRITE MPS QUOTED NAMES, 314
MSK IPAR WRITE MPS STRICT, 310
MSK IPAR WRITE PRECISION, 319
MSK IPAR WRITE SOL CONSTRAINTS, 313
MSK IPAR WRITE SOL HEAD, 323
MSK IPAR WRITE SOL VARIABLES, 321
MSK IPAR WRITE TASK INC SOL, 317
MSK IPAR WRITE XML MODE, 320
MSK LANG DAN, 326
MSK LANG ENG, 326
MSK LICENSE BUFFER LENGTH, 361
MSK LIINF BI CLEAN DUAL DEG ITER, 327
MSK LIINF BI CLEAN DUAL ITER, 327
MSK LIINF BI CLEAN PRIMAL DEG ITER, 327
MSK LIINF BI CLEAN PRIMAL DUAL DEG ITER, 326
MSK LIINF BI CLEAN PRIMAL DUAL ITER, 326
MSK LIINF BI CLEAN PRIMAL DUAL SUB ITER, 327
MSK LIINF BI CLEAN PRIMAL ITER, 326
MSK LIINF BI DUAL ITER, 327
MSK LIINF BI PRIMAL ITER, 327
MSK LIINF INTPNT FACTOR NUM NZ, 326
MSK LIINF MIO INTPNT ITER, 326
MSK LIINF MIO SIMPLEX ITER, 327
MSK LIINF RD NUMANZ, 327
MSK LIINF RD NUMQNZ, 327
MSK MARK LO, 327
MSK MARK UP, 327
MSK MAX STR LEN, 361
MSK MIO CONT SOL ITG, 327
MSK MIO CONT SOL ITG REL, 328
MSK MIO CONT SOL NONE, 328
MSK MIO CONT SOL ROOT, 328
MSK MIO MODE IGNORED, 328
MSK MIO MODE LAZY, 328
MSK MIO MODE SATISFIED, 328
MSK MIO NODE SELECTION BEST, 328
MSK MIO NODE SELECTION FIRST, 329
MSK MIO NODE SELECTION FREE, 328
MSK MIO NODE SELECTION HYBRID, 328
MSK MIO NODE SELECTION PSEUDO, 328
MSK MIO NODE SELECTION WORST, 328
MSK MPS FORMAT FREE, 329
MSK MPS FORMAT RELAXED, 329
MSK MPS FORMAT STRICT, 329
MSK MSG MPS SELECTED, 329
MSK MSG READING FILE, 329
MSK MSG WRITING FILE, 329
MSK NETWORK DETECT ADVANCED, 329
MSK NETWORK DETECT FREE, 329
MSK NETWORK DETECT SIMPLE, 329
MSK OBJECTIVE SENSE MAXIMIZE, 330
MSK OBJECTIVE SENSE MINIMIZE, 330
MSK OBJECTIVE SENSE UNDEFINED, 330
MSK OFF, 330
MSK ON, 330
MSK OPTIMIZER CONCURRENT, 330
MSK OPTIMIZER CONIC, 330
MSK OPTIMIZER DUAL SIMPLEX, 330
382 INDEX
MSK OPTIMIZER FREE, 330
MSK OPTIMIZER FREE SIMPLEX, 331
MSK OPTIMIZER INTPNT, 330
MSK OPTIMIZER MIXED INT, 330
MSK OPTIMIZER NONCONVEX, 330
MSK OPTIMIZER PRIMAL DUAL SIMPLEX, 330
MSK OPTIMIZER PRIMAL SIMPLEX, 331
MSK OPTIMIZER QCONE, 330
MSK ORDER METHOD APPMINLOC1, 331
MSK ORDER METHOD APPMINLOC2, 331
MSK ORDER METHOD FREE, 331
MSK ORDER METHOD GRAPHPAR1, 331
MSK ORDER METHOD GRAPHPAR2, 331
MSK ORDER METHOD NONE, 331
MSK PAR DOU TYPE, 331
MSK PAR INT TYPE, 331
MSK PAR INVALID TYPE, 331
MSK PAR STR TYPE, 331
MSK PI CON, 332
MSK PI CONE, 332
MSK PI VAR, 332
MSK PRESOLVE MODE FREE, 332
MSK PRESOLVE MODE OFF, 332
MSK PRESOLVE MODE ON, 332
MSK PRO STA DUAL FEAS, 333
MSK PRO STA DUAL INFEAS, 333
MSK PRO STA ILL POSED, 333
MSK PRO STA NEAR DUAL FEAS, 333
MSK PRO STA NEAR PRIM AND DUAL FEAS, 333
MSK PRO STA NEAR PRIM FEAS, 333
MSK PRO STA PRIM AND DUAL FEAS, 333
MSK PRO STA PRIM AND DUAL INFEAS, 333
MSK PRO STA PRIM FEAS, 333
MSK PRO STA PRIM INFEAS, 333
MSK PRO STA PRIM INFEAS OR UNBOUNDED, 333
MSK PRO STA UNKNOWN, 333
MSK PROBTYPE CONIC, 332
MSK PROBTYPE GECO, 332
MSK PROBTYPE LO, 332
MSK PROBTYPE MIXED, 332
MSK PROBTYPE QCQO, 332
MSK PROBTYPE QO, 332
MSK Q READ ADD, 333
MSK Q READ DROP LOWER, 333
MSK Q READ DROP UPPER, 333
MSK RES ERR AD INVALID CODELIST, 341
MSK RES ERR AD INVALID OPERAND, 349
MSK RES ERR AD INVALID OPERATOR, 352
MSK RES ERR AD MISSING OPERAND, 348
MSK RES ERR AD MISSING RETURN, 343
MSK RES ERR API ARRAY TOO SMALL, 342
MSK RES ERR API CB CONNECT, 347
MSK RES ERR API FATAL ERROR, 341
MSK RES ERR API INTERNAL, 344
MSK RES ERR ARGUMENT DIMENSION, 348
MSK RES ERR ARGUMENT LENNEQ, 346
MSK RES ERR ARGUMENT PERM ARRAY, 337
MSK RES ERR ARGUMENT TYPE, 337
MSK RES ERR BASIS, 350
MSK RES ERR BASIS FACTOR, 336
MSK RES ERR BASIS SINGULAR, 347
MSK RES ERR BLANK NAME, 346
MSK RES ERR CANNOT CLONE NL, 334
MSK RES ERR CANNOT HANDLE NL, 341
MSK RES ERR CON Q NOT NSD, 344
MSK RES ERR CON Q NOT PSD, 338
MSK RES ERR CONCURRENT OPTIMIZER, 346
MSK RES ERR CONE INDEX, 339
MSK RES ERR CONE OVERLAP, 345
MSK RES ERR CONE REP VAR, 341
MSK RES ERR CONE SIZE, 351
MSK RES ERR CONE TYPE, 345
MSK RES ERR CONE TYPE STR, 351
MSK RES ERR DATA FILE EXT, 337
MSK RES ERR DUP NAME, 346
MSK RES ERR END OF FILE, 337
MSK RES ERR FACTOR, 338
MSK RES ERR FEASREPAIR CANNOT RELAX, 336
MSK RES ERR FEASREPAIR INCONSISTENT BOUND, 344
MSK RES ERR FEASREPAIR SOLVING RELAXED, 342
MSK RES ERR FILE LICENSE, 347
MSK RES ERR FILE OPEN, 341
MSK RES ERR FILE READ, 342
MSK RES ERR FILE WRITE, 340
MSK RES ERR FIRST, 351
MSK RES ERR FIRSTI, 337
MSK RES ERR FIRSTJ, 337
MSK RES ERR FIXED BOUND VALUES, 344
MSK RES ERR FLEXLM, 351
MSK RES ERR HUGE AIJ, 348
MSK RES ERR HUGE C, 352
MSK RES ERR IDENTICAL TASKS, 345
MSK RES ERR IN ARGUMENT, 346
MSK RES ERR INDEX, 335
MSK RES ERR INDEX ARR IS TOO LARGE, 351
MSK RES ERR INDEX ARR IS TOO SMALL, 342
MSK RES ERR INDEX IS TOO LARGE, 336
MSK RES ERR INDEX IS TOO SMALL, 334
MSK RES ERR INF DOU INDEX, 337
MSK RES ERR INF DOU NAME, 349
INDEX 383
MSK RES ERR INF INT INDEX, 342
MSK RES ERR INF INT NAME, 344
MSK RES ERR INF LINT INDEX, 336
MSK RES ERR INF LINT NAME, 335
MSK RES ERR INF TYPE, 340
MSK RES ERR INFEAS UNDEFINED, 343
MSK RES ERR INFINITE BOUND, 342
MSK RES ERR INT64 TO INT32 CAST, 352
MSK RES ERR INTERNAL, 336
MSK RES ERR INTERNAL TEST FAILED, 346
MSK RES ERR INV APTRE, 347
MSK RES ERR INV BK, 351
MSK RES ERR INV BKC, 345
MSK RES ERR INV BKX, 350
MSK RES ERR INV CONE TYPE, 349
MSK RES ERR INV CONE TYPE STR, 348
MSK RES ERR INV MARKI, 334
MSK RES ERR INV MARKJ, 352
MSK RES ERR INV NAME ITEM, 343
MSK RES ERR INV NUMI, 338
MSK RES ERR INV NUMJ, 338
MSK RES ERR INV OPTIMIZER, 341
MSK RES ERR INV PROBLEM, 334
MSK RES ERR INV QCON SUBI, 351
MSK RES ERR INV QCON SUBJ, 337
MSK RES ERR INV QCON SUBK, 337
MSK RES ERR INV QCON VAL, 339
MSK RES ERR INV QOBJ SUBI, 345
MSK RES ERR INV QOBJ SUBJ, 345
MSK RES ERR INV QOBJ VAL, 342
MSK RES ERR INV SK, 349
MSK RES ERR INV SK STR, 349
MSK RES ERR INV SKC, 335
MSK RES ERR INV SKN, 335
MSK RES ERR INV SKX, 334
MSK RES ERR INV VAR TYPE, 351
MSK RES ERR INVALID ACCMODE, 343
MSK RES ERR INVALID AMPL STUB, 348
MSK RES ERR INVALID BRANCH DIRECTION, 345
MSK RES ERR INVALID BRANCH PRIORITY, 338
MSK RES ERR INVALID COMPRESSION, 346
MSK RES ERR INVALID FILE NAME, 340
MSK RES ERR INVALID FORMAT TYPE, 348
MSK RES ERR INVALID IOMODE, 343
MSK RES ERR INVALID MBT FILE, 344
MSK RES ERR INVALID NAME IN SOL FILE, 334
MSK RES ERR INVALID OBJ NAME, 341
MSK RES ERR INVALID OBJECTIVE SENSE, 347
MSK RES ERR INVALID SOL FILE NAME, 348
MSK RES ERR INVALID STREAM, 343
MSK RES ERR INVALID SURPLUS, 351
MSK RES ERR INVALID TASK, 341
MSK RES ERR INVALID UTF8, 352
MSK RES ERR INVALID WCHAR, 337
MSK RES ERR LAST, 342
MSK RES ERR LASTI, 337
MSK RES ERR LASTJ, 335
MSK RES ERR LICENSE, 337
MSK RES ERR LICENSE CANNOT ALLOCATE, 345
MSK RES ERR LICENSE CANNOT CONNECT, 351
MSK RES ERR LICENSE EXPIRED, 335
MSK RES ERR LICENSE FEATURE, 341
MSK RES ERR LICENSE INVALID HOSTID, 343
MSK RES ERR LICENSE MAX, 344
MSK RES ERR LICENSE MOSEKLM DAEMON, 337
MSK RES ERR LICENSE NO SERVER SUPPORT, 339
MSK RES ERR LICENSE SERVER, 350
MSK RES ERR LICENSE SERVER VERSION, 343
MSK RES ERR LICENSE VERSION, 338
MSK RES ERR LINK FILE DLL, 342
MSK RES ERR LIVING TASKS, 339
MSK RES ERR LP DUP SLACK NAME, 338
MSK RES ERR LP EMPTY, 343
MSK RES ERR LP FILE FORMAT, 351
MSK RES ERR LP FORMAT, 347
MSK RES ERR LP FREE CONSTRAINT, 347
MSK RES ERR LP INCOMPATIBLE, 336
MSK RES ERR LP INVALID CON NAME, 340
MSK RES ERR LP INVALID VAR NAME, 336
MSK RES ERR LP WRITE CONIC PROBLEM, 338
MSK RES ERR LP WRITE GECO PROBLEM, 347
MSK RES ERR LU MAX NUM TRIES, 335
MSK RES ERR MAXNUMCON, 338
MSK RES ERR MAXNUMCONE, 339
MSK RES ERR MAXNUMQNZ, 340
MSK RES ERR MAXNUMVAR, 346
MSK RES ERR MBT INCOMPATIBLE, 343
MSK RES ERR MIO NO OPTIMIZER, 334
MSK RES ERR MIO NOT LOADED, 345
MSK RES ERR MISSING LICENSE FILE, 335
MSK RES ERR MIXED PROBLEM, 336
MSK RES ERR MPS CONE OVERLAP, 347
MSK RES ERR MPS CONE REPEAT, 348
MSK RES ERR MPS CONE TYPE, 351
MSK RES ERR MPS FILE, 338
MSK RES ERR MPS INV BOUND KEY, 350
MSK RES ERR MPS INV CON KEY, 344
MSK RES ERR MPS INV FIELD, 346
MSK RES ERR MPS INV MARKER, 349
MSK RES ERR MPS INV SEC NAME, 350
384 INDEX
MSK RES ERR MPS INV SEC ORDER, 344
MSK RES ERR MPS INVALID OBJ NAME, 350
MSK RES ERR MPS INVALID OBJSENSE, 346
MSK RES ERR MPS MUL CON NAME, 343
MSK RES ERR MPS MUL CSEC, 346
MSK RES ERR MPS MUL QOBJ, 334
MSK RES ERR MPS MUL QSEC, 349
MSK RES ERR MPS NO OBJECTIVE, 342
MSK RES ERR MPS NULL CON NAME, 340
MSK RES ERR MPS NULL VAR NAME, 343
MSK RES ERR MPS SPLITTED VAR, 348
MSK RES ERR MPS TAB IN FIELD2, 340
MSK RES ERR MPS UNDEF CON NAME, 352
MSK RES ERR MPS UNDEF VAR NAME, 336
MSK RES ERR MUL A ELEMENT, 334
MSK RES ERR NAME IS NULL, 351
MSK RES ERR NAME MAX LEN, 345
MSK RES ERR NAN IN AIJ, 351
MSK RES ERR NAN IN BLC, 339
MSK RES ERR NAN IN BLX, 349
MSK RES ERR NAN IN BUC, 337
MSK RES ERR NAN IN BUX, 350
MSK RES ERR NAN IN C, 339
MSK RES ERR NAN IN DOUBLE DATA, 350
MSK RES ERR NEGATIVE APPEND, 349
MSK RES ERR NEGATIVE SURPLUS, 336
MSK RES ERR NEWER DLL, 349
MSK RES ERR NO BASIS SOL, 341
MSK RES ERR NO DUAL FOR ITG SOL, 336
MSK RES ERR NO DUAL INFEAS CER, 334
MSK RES ERR NO INIT ENV, 334
MSK RES ERR NO OPTIMIZER VAR TYPE, 352
MSK RES ERR NO PRIMAL INFEAS CER, 339
MSK RES ERR NO SOLUTION IN CALLBACK, 352
MSK RES ERR NONCONVEX, 347
MSK RES ERR NONLINEAR EQUALITY, 337
MSK RES ERR NONLINEAR RANGED, 338
MSK RES ERR NR ARGUMENTS, 338
MSK RES ERR NULL ENV, 346
MSK RES ERR NULL POINTER, 341
MSK RES ERR NULL TASK, 345
MSK RES ERR NUMCONLIM, 343
MSK RES ERR NUMVARLIM, 349
MSK RES ERR OBJ Q NOT NSD, 341
MSK RES ERR OBJ Q NOT PSD, 335
MSK RES ERR OBJECTIVE RANGE, 349
MSK RES ERR OLDER DLL, 348
MSK RES ERR OPEN DL, 342
MSK RES ERR OPF FORMAT, 346
MSK RES ERR OPF NEW VARIABLE, 351
MSK RES ERR OPF PREMATURE EOF, 339
MSK RES ERR OPTIMIZER LICENSE, 347
MSK RES ERR ORD INVALID, 341
MSK RES ERR ORD INVALID BRANCH DIR, 352
MSK RES ERR OVERFLOW, 336
MSK RES ERR PARAM INDEX, 337
MSK RES ERR PARAM IS TOO LARGE, 347
MSK RES ERR PARAM IS TOO SMALL, 338
MSK RES ERR PARAM NAME, 336
MSK RES ERR PARAM NAME DOU, 339
MSK RES ERR PARAM NAME INT, 336
MSK RES ERR PARAM NAME STR, 341
MSK RES ERR PARAM TYPE, 334
MSK RES ERR PARAM VALUE STR, 350
MSK RES ERR PLATFORM NOT LICENSED, 348
MSK RES ERR POSTSOLVE, 347
MSK RES ERR PRO ITEM, 347
MSK RES ERR PROB LICENSE, 349
MSK RES ERR QCON SUBI TOO LARGE, 352
MSK RES ERR QCON SUBI TOO SMALL, 347
MSK RES ERR QCON UPPER TRIANGLE, 340
MSK RES ERR QOBJ UPPER TRIANGLE, 340
MSK RES ERR READ FORMAT, 347
MSK RES ERR READ LP MISSING END TAG, 348
MSK RES ERR READ LP NONEXISTING NAME, 338
MSK RES ERR REMOVE CONE VARIABLE, 341
MSK RES ERR SEN BOUND INVALID LO, 348
MSK RES ERR SEN BOUND INVALID UP, 344
MSK RES ERR SEN FORMAT, 339
MSK RES ERR SEN INDEX INVALID, 335
MSK RES ERR SEN INDEX RANGE, 339
MSK RES ERR SEN INVALID REGEXP, 344
MSK RES ERR SEN NUMERICAL, 339
MSK RES ERR SEN SOLUTION STATUS, 336
MSK RES ERR SEN UNDEF NAME, 348
MSK RES ERR SIZE LICENSE, 352
MSK RES ERR SIZE LICENSE CON, 347
MSK RES ERR SIZE LICENSE INTVAR, 352
MSK RES ERR SIZE LICENSE NUMCORES, 338
MSK RES ERR SIZE LICENSE VAR, 343
MSK RES ERR SOL FILE INVALID NUMBER, 335
MSK RES ERR SOLITEM, 347
MSK RES ERR SOLVER PROBTYPE, 342
MSK RES ERR SPACE, 346
MSK RES ERR SPACE LEAKING, 348
MSK RES ERR SPACE NO INFO, 336
MSK RES ERR THREAD COND INIT, 349
MSK RES ERR THREAD CREATE, 340
INDEX 385
MSK RES ERR THREAD MUTEX INIT, 343
MSK RES ERR THREAD MUTEX LOCK, 342
MSK RES ERR THREAD MUTEX UNLOCK, 338
MSK RES ERR TOO SMALL MAXNUMANZ, 346
MSK RES ERR UNB STEP SIZE, 342
MSK RES ERR UNDEF SOLUTION, 334
MSK RES ERR UNDEFINED OBJECTIVE SENSE, 352
MSK RES ERR UNKNOWN, 338
MSK RES ERR USER FUNC RET, 345
MSK RES ERR USER FUNC RET DATA, 337
MSK RES ERR USER NLO EVAL, 340
MSK RES ERR USER NLO EVAL HESSUBI, 342
MSK RES ERR USER NLO EVAL HESSUBJ, 342
MSK RES ERR USER NLO FUNC, 337
MSK RES ERR WHICHITEM NOT ALLOWED, 349
MSK RES ERR WHICHSOL, 349
MSK RES ERR WRITE LP FORMAT, 339
MSK RES ERR WRITE LP NON UNIQUE NAME, 350
MSK RES ERR WRITE MPS INVALID NAME, 345
MSK RES ERR WRITE OPF INVALID VAR NAME, 338
MSK RES ERR WRITING FILE, 352
MSK RES ERR XML INVALID PROBLEM TYPE, 344
MSK RES ERR Y IS UNDEFINED, 344
MSK RES OK, 347
MSK RES TRM INTERNAL, 342
MSK RES TRM INTERNAL STOP, 351
MSK RES TRM MAX ITERATIONS, 352
MSK RES TRM MAX NUM SETBACKS, 345
MSK RES TRM MAX TIME, 348
MSK RES TRM MIO NEAR ABS GAP, 334
MSK RES TRM MIO NEAR REL GAP, 349
MSK RES TRM MIO NUM BRANCHES, 334
MSK RES TRM MIO NUM RELAXS, 345
MSK RES TRM NUM MAX NUM INT SOLUTIONS, 345
MSK RES TRM NUMERICAL PROBLEM, 344
MSK RES TRM OBJECTIVE RANGE, 351
MSK RES TRM STALL, 350
MSK RES TRM USER BREAK, 350
MSK RES TRM USER CALLBACK, 344
MSK RES WRN ANA ALMOST INT BOUNDS, 345
MSK RES WRN ANA C ZERO, 343
MSK RES WRN ANA CLOSE BOUNDS, 335
MSK RES WRN ANA EMPTY COLS, 350
MSK RES WRN ANA LARGE BOUNDS, 346
MSK RES WRN CONSTRUCT INVALID SOL ITG, 345
MSK RES WRN CONSTRUCT NO SOL ITG, 348
MSK RES WRN CONSTRUCT SOLUTION INFEAS, 344
MSK RES WRN DROPPED NZ QOBJ, 336
MSK RES WRN ELIMINATOR SPACE, 349
MSK RES WRN EMPTY NAME, 349
MSK RES WRN IGNORE INTEGER, 341
MSK RES WRN INCOMPLETE LINEAR DEPENDENCY CHECK,
340
MSK RES WRN LARGE AIJ, 341
MSK RES WRN LARGE BOUND, 346
MSK RES WRN LARGE CJ, 348
MSK RES WRN LARGE CON FX, 343
MSK RES WRN LARGE LO BOUND, 344
MSK RES WRN LARGE UP BOUND, 343
MSK RES WRN LICENSE EXPIRE, 346
MSK RES WRN LICENSE FEATURE EXPIRE, 336
MSK RES WRN LICENSE SERVER, 346
MSK RES WRN LP DROP VARIABLE, 337
MSK RES WRN LP OLD QUAD FORMAT, 349
MSK RES WRN MIO INFEASIBLE FINAL, 340
MSK RES WRN MPS SPLIT BOU VECTOR, 343
MSK RES WRN MPS SPLIT RAN VECTOR, 348
MSK RES WRN MPS SPLIT RHS VECTOR, 344
MSK RES WRN NAME MAX LEN, 351
MSK RES WRN NO GLOBAL OPTIMIZER, 341
MSK RES WRN NZ IN UPR TRI, 336
MSK RES WRN OPEN PARAM FILE, 347
MSK RES WRN PRESOLVE BAD PRECISION, 334
MSK RES WRN PRESOLVE OUTOFSPACE, 352
MSK RES WRN SOL FILE IGNORED CON, 350
MSK RES WRN SOL FILE IGNORED VAR, 334
MSK RES WRN SOL FILTER, 342
MSK RES WRN SPAR MAX LEN, 339
MSK RES WRN TOO FEW BASIS VARS, 350
MSK RES WRN TOO MANY BASIS VARS, 336
MSK RES WRN UNDEF SOL FILE NAME, 351
MSK RES WRN USING GENERIC NAMES, 340
MSK RES WRN WRITE DISCARDED CFIX, 340
MSK RES WRN ZERO AIJ, 338
MSK RES WRN ZEROS IN SPARSE COL, 340
MSK RES WRN ZEROS IN SPARSE ROW, 337
MSK RESPONSE ERR, 353
MSK RESPONSE OK, 353
MSK RESPONSE TRM, 352
MSK RESPONSE UNK, 352
MSK RESPONSE WRN, 352
MSK SCALING AGGRESSIVE, 353
MSK SCALING FREE, 353
MSK SCALING METHOD FREE, 353
MSK SCALING METHOD POW2, 353
MSK SCALING MODERATE, 353
MSK SCALING NONE, 353
MSK SENSITIVITY TYPE BASIS, 353
MSK SENSITIVITY TYPE OPTIMAL PARTITION, 353
MSK SIM DEGEN AGGRESSIVE, 354
386 INDEX
MSK SIM DEGEN FREE, 354
MSK SIM DEGEN MINIMUM, 354
MSK SIM DEGEN MODERATE, 354
MSK SIM DEGEN NONE, 354
MSK SIM EXPLOIT DUPVEC FREE, 354
MSK SIM EXPLOIT DUPVEC OFF, 354
MSK SIM EXPLOIT DUPVEC ON, 354
MSK SIM HOTSTART FREE, 354
MSK SIM HOTSTART NONE, 354
MSK SIM HOTSTART STATUS KEYS, 354
MSK SIM REFORMULATION AGGRESSIVE, 355
MSK SIM REFORMULATION FREE, 355
MSK SIM REFORMULATION OFF, 355
MSK SIM REFORMULATION ON, 354
MSK SIM SELECTION ASE, 355
MSK SIM SELECTION DEVEX, 355
MSK SIM SELECTION FREE, 355
MSK SIM SELECTION FULL, 355
MSK SIM SELECTION PARTIAL, 355
MSK SIM SELECTION SE, 355
MSK SK BAS, 359
MSK SK FIX, 359
MSK SK INF, 359
MSK SK LOW, 359
MSK SK SUPBAS, 359
MSK SK UNK, 359
MSK SK UPR, 360
MSK SOL BAS, 357
MSK SOL ITEM SLC, 356
MSK SOL ITEM SLX, 356
MSK SOL ITEM SNX, 356
MSK SOL ITEM SUC, 355
MSK SOL ITEM SUX, 356
MSK SOL ITEM XC, 355
MSK SOL ITEM XX, 355
MSK SOL ITEM Y, 355
MSK SOL ITG, 357
MSK SOL ITR, 357
MSK SOL STA DUAL FEAS, 357
MSK SOL STA DUAL INFEAS CER, 356
MSK SOL STA INTEGER OPTIMAL, 356
MSK SOL STA NEAR DUAL FEAS, 356
MSK SOL STA NEAR DUAL INFEAS CER, 356
MSK SOL STA NEAR INTEGER OPTIMAL, 356
MSK SOL STA NEAR OPTIMAL, 356
MSK SOL STA NEAR PRIM AND DUAL FEAS, 356
MSK SOL STA NEAR PRIM FEAS, 356
MSK SOL STA NEAR PRIM INFEAS CER, 356
MSK SOL STA OPTIMAL, 356
MSK SOL STA PRIM AND DUAL FEAS, 356
MSK SOL STA PRIM FEAS, 356
MSK SOL STA PRIM INFEAS CER, 356
MSK SOL STA UNKNOWN, 356
MSK SOLVE DUAL, 357
MSK SOLVE FREE, 357
MSK SOLVE PRIMAL, 357
MSK SPAR BAS SOL FILE NAME, 357
MSK SPAR DATA FILE NAME, 358
MSK SPAR DEBUG FILE NAME, 359
MSK SPAR FEASREPAIR NAME PREFIX, 357
MSK SPAR FEASREPAIR NAME SEPARATOR, 358
MSK SPAR FEASREPAIR NAME WSUMVIOL, 358
MSK SPAR INT SOL FILE NAME, 358
MSK SPAR ITR SOL FILE NAME, 359
MSK SPAR PARAM COMMENT SIGN, 357
MSK SPAR PARAM READ FILE NAME, 359
MSK SPAR PARAM WRITE FILE NAME, 358
MSK SPAR READ MPS BOU NAME, 358
MSK SPAR READ MPS OBJ NAME, 357
MSK SPAR READ MPS RAN NAME, 358
MSK SPAR READ MPS RHS NAME, 358
MSK SPAR SENSITIVITY FILE NAME, 359
MSK SPAR SENSITIVITY RES FILE NAME, 359
MSK SPAR SOL FILTER XC LOW, 358
MSK SPAR SOL FILTER XC UPR, 358
MSK SPAR SOL FILTER XX LOW, 359
MSK SPAR SOL FILTER XX UPR, 358
MSK SPAR STAT FILE NAME, 358
MSK SPAR STAT KEY, 359
MSK SPAR STAT NAME, 359
MSK SPAR WRITE LP GEN VAR NAME, 358
MSK STARTING POINT CONSTANT, 360
MSK STARTING POINT FREE, 360
MSK STARTING POINT GUESS, 360
MSK STARTING POINT SATISFY BOUNDS, 360
MSK STREAM ERR, 360
MSK STREAM LOG, 360
MSK STREAM MSG, 360
MSK STREAM WRN, 360
MSK VAR TYPE CONT, 361
MSK VAR TYPE INT, 361
MSK WRITE XML MODE COL, 361
MSK WRITE XML MODE ROW, 361
timing level (parameter), 264
upper obj cut (parameter), 196
upper obj cut finite trh (parameter), 196
variables
decision, 35, 48, 109
INDEX 387
lower limit, 36, 48, 109
upper limit, 36, 48, 109
warning level (parameter), 264
write bas constraints (parameter), 264
write bas head (parameter), 265
write bas variables (parameter), 265
write data compressed (parameter), 265
write data format (parameter), 265
write data param (parameter), 266
write free con (parameter), 266
write generic names (parameter), 266
write generic names io (parameter), 267
write int constraints (parameter), 267
write int head (parameter), 267
write int variables (parameter), 267
write lp gen var name (parameter), 279
write lp line width (parameter), 268
write lp quoted names (parameter), 268
write lp strict format (parameter), 268
write lp terms per line (parameter), 268
write mps int (parameter), 269
write mps obj sense (parameter), 269
write mps quoted names (parameter), 269
write mps strict (parameter), 270
write precision (parameter), 270
write sol constraints (parameter), 270
write sol head (parameter), 270
write sol variables (parameter), 271
write task inc sol (parameter), 271
write xml mode (parameter), 271
xml format, 141

MOSEK

Uploaded by

Copyright:

Available Formats

MOSEK

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MOSEK

Uploaded by

Copyright:

Available Formats

The MOSEK optimization tools

is a product of the cones

x . First dene the set

> 0. It follows that

is a primal optimal solution and (

) is a dual optimal solution; this is reported

is a certicate of dual infeasibility, whereas if (10.6) is satised

is a certicate of dual infeasibility.

denote the reported primal solution. MOSEK states

is a certicate of dual infeasibility then for any j such that

is an optimal objective value to (13.4). An extension of the idea presented above

) is an arbitrary optimal solution then it must hold

is an arbitrary primal optimal solution then it must hold

is the value for the x solution.

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.