Decomposing 1-Sperner hypergraphs

Endre Boros
MSIS Department and RUTCOR, Rutgers University, New Jersey, USA
100 Rockafeller Rd, Piscataway NJ 08854, USA
Endre.Boros@rutgers.edu

Vladimir Gurvich∗
National Research University: Higher School of Economics, Moscow, Russia
vgurvich@hse.ru

Martin Milanič†
University of Primorska, UP IAM, Muzejski trg 2, SI-6000 Koper, Slovenia
University of Primorska, UP FAMNIT, Glagoljaška 8, SI-6000 Koper, Slovenia
martin.milanic@upr.si

Submitted: May 28, 2018; Accepted: Jul 4, 2019; Published: Jul 19, 2019
c The authors. Released under the CC BY-ND license (International 4.0).

Abstract
A hypergraph is Sperner if no hyperedge contains another one. A Sperner hy-
pergraph is equilizable (resp., threshold ) if the characteristic vectors of its hyper-
edges are the (minimal) binary solutions to a linear equation (resp., inequality)
with positive coefficients. These combinatorial notions have many applications and
are motivated by the theory of Boolean functions and integer programming. We
introduce in this paper the class of 1-Sperner hypergraphs, defined by the property
that for every two hyperedges the smallest of their two set differences is of size
one. We characterize this class of Sperner hypergraphs by a decomposition theo-
rem and derive several consequences from it. In particular, we obtain bounds on
the size of 1-Sperner hypergraphs and their transversal hypergraphs, show that the
characteristic vectors of the hyperedges are linearly independent over the reals, and
prove that 1-Sperner hypergraphs are both threshold and equilizable. The study
of 1-Sperner hypergraphs is motivated also by their applications in graph theory,
which we present in a companion paper.
Mathematics Subject Classifications: 05C65
∗
Partially funded by the Russian Academic Excellence Project ‘5-100’.
†
Supported in part by the Slovenian Research Agency (I0-0035, research program P1-0285 and research
projects J1-9110, N1-0102).

the electronic journal of combinatorics 26(3) (2019), #P3.18 1

1 Introduction
In this paper we consider various classes of hypergraphs, with a focus on the newly
introduced class of 1-Sperner hypergraphs. As we will see, this is an interesting and useful
notion with many surprising properties, including a simple recursive structure. Before
we explain and motivate our study and results, we overview the necessary background
definitions.

1.1 Background
A hypergraph H is a pair (V, E) where V = V (H) is a finite set of vertices and E = E(H)
is a set of subsets of V , called hyperedges [6]. Given a positive integer k, a hypergraph H
is said to be k-uniform if |e| = k for all e ∈ E(H), and uniform if it is k-uniform for some
k. In particular, the (finite, simple, and undirected) graphs are precisely the 2-uniform
hypergraphs. Four properties of hypergraphs will be particularly relevant for our study:
Sperner, threshold, equilizable, and dually Sperner hypergraphs.

Sperner hypergraphs. A hypergraph is said to be Sperner if no hyperedge contains

another one, that is, if e, f ∈ E and e ⊆ f implies e = f ; see, e.g., [5, 40, 42]. Sperner
hypergraphs were studied in the literature under different names including simple hyper-
graphs by Berge [6], clutters by Billera [7, 8] and by Edmonds and Fulkerson [19, 22], and
coalitions in the game theory literature [43]. See also [26] for additional references on
applications of Sperner hypergraphs in other areas of mathematics.

Threshold hypergraphs. A hypergraph H = (V, E) is said to be threshold if there exist

a non-negative integer weight function w : V → Z>0 and a non-negative
P integer threshold
t ∈ Z>0 such that for every subset X ⊆ V , we have w(X) := x∈X w(x) > t if and only if
e ⊆ X for some e ∈ E. A pair (w, t) as above will be referred to as a threshold separator
of H. Note that the classes of Sperner and threshold hypergraphs are incomparable (this
issue will be discussed in more detail in Sections 1.4 and 2.3). Furthermore, the mapping
that takes every subset U ⊆ V to its characteristic vector χU ∈ {0, 1}V , defined by

U 1, if v ∈ U
χv =
0, otherwise ,

shows that the sets of hyperedges of threshold Sperner hypergraphs are in a one-to-one
correspondence with the sets of minimal feasible binary solutions of the linear inequality
w> x > t. A set of vertices X ⊆ V in a hypergraph is said to be independent (or stable) if it
does not contain any hyperedge, and dependent otherwise. Thus, threshold hypergraphs
are exactly the hypergraphs admitting a linear function on the vertices separating the
characteristic vectors of the independent sets from the characteristic vectors of dependent
sets.
Threshold hypergraphs were defined in the uniform case by Golumbic [24] and studied
further by Reiterman et al. [38]. The 2-uniform threshold hypergraphs are precisely
the threshold graphs, introduced by Chvátal and Hammer [16] and studied afterwards

the electronic journal of combinatorics 26(3) (2019), #P3.18 2

in numerous papers; see also the monograph by Mahadev and Peled [28]. In their full
generality (that is, without the restriction that the hypergraph is uniform), the concept
of threshold hypergraphs is equivalent to that of threshold monotone Boolean functions.
Threshold Boolean functions provide a simple but fundamental model for many questions
investigated in a variety of areas including electrical engineering, artificial intelligence,
game theory, cryptography, and many others; see, e.g., [3, 17, 18, 33].
Close interrelations between hypergraphs and monotone Boolean functions are often
useful in the study of threshold and other hypergraphs, allowing for the transfer and
applications of results from the theory of Boolean functions; see [18]. For example, a
polynomial-time recognition algorithm for threshold monotone Boolean functions repre-
sented by their complete disjunctive normal form was given by Peled and Simeone [37].
The algorithm is based on linear programming and implies a polynomial-time recognition
algorithm for threshold hypergraphs. To the best of our knowledge, no ‘purely combina-
torial’ polynomial-time recognition algorithm for threshold hypergraphs is known [18].1

Equilizable hypergraphs. Replacing a linear inequality with positive coefficients by

a linear equation maps the notion of threshold hypegraphs to the notion of equilizable
hypergraphs. A hypergraph H = (V, E) is said to be equilizable if there exist a (strictly)
positive integer weight function w : V → Z>0 and a non-negative integer threshold t ∈ Z>0
such that for every subset X ⊆ V , we have w(X) = t if and only if X ∈ E.
Depending on the context, one may want to relax the assumption that all the weights
are strictly positive to allow zero weights. However, we find the assumption of strictly
positive weights useful for our study; in particular, it implies the following.

Proposition 1. Every equilizable hypergraph is Sperner.

Equilizable hypergraphs are a very natural family. The sets of hyperedges of an equil-
izable hypergraph are in a one-to-one correspondence with the sets of binary solutions to
a linear equality of the form w> x = t where w ∈ ZV>0 and t ∈ Z>0 , that is, with the sets of
binary vectors that could be cut out by a single hyperplane with positive coefficients from
the unit hypercube. So, it is not surprising that properties of equilizable hypergraphs are
fundamental in integer (or binary) programming. In particular, an old result of Math-
ews [30] shows how to reduce two linear Diophantine equations in non-negative integers
with strictly positive coefficients to a single equivalent linear equation of the same type.
Motivated by integer programming considerations, many authors generalized Mathews’
result in a variety of ways, see [2,11,21,23,34,39].2 Furthermore, Mathews’ result implies
1
See Smaus [41] for an attempt.
2
In some sense, these equation aggregation results are not unexpected. The intersection of two hyper-
planes in Rn is an (n − 2)-dimensional subspace F. Any integer point not included in F extends F to a
unique hyperplane (single equality). Since in a bounded region there are only finitely many such feasible
integer points, there are only finitely many hyperplanes through F that contain an integer point not con-
tained in F. Thus we must have infinitely many hyperplanes containing F that do not contain any other
integer feasible point. The only nontrivial part is the numerical construction of an explicit hyperplane.
All constructions from [2, 11, 21, 23, 30, 34, 39] end up introducing exponentially growing coefficients.

the electronic journal of combinatorics 26(3) (2019), #P3.18 3

that the class of equilizable hypergraphs on a given vertex set is closed under intersec-
tion.3 A related question about linear inequalities led Chvátal and Hammer [16] to the
introduction of threshold graphs.
Equilizable hypergraphs can also be seen as a generalization of the class of equistable
graphs, defined as follows. A stable set (or: independent set) in a graph G is a set of
pairwise non-adjacent vertices. A graph G = (V, E) is said to be equistable if there exist
a (strictly) positive integer weight function w : V → Z>0 and a non-negative integer
threshold t such that for every subset X ⊆ V , we have w(X) = t if and only if X is
an (inclusion-)maximal stable set of G. The connection between equistable graphs and
equilizable hypergraphs can be easily explained using the notion of stable set hypergraphs.
The stable set hypergraph of a graph G is the hypergraph S(G) with vertex set V (G) and
in which the hyperedges are exactly the maximal stable sets of G. Clearly, a graph G is
equistable if and only if its stable set hypergraph is equilizable. Equistable graphs were
introduced in 1980 by Payan [35], who proved that every threshold graph is equistable.
While equistable graphs were originally defined using a function ϕ : V → R>0 such that
for every subset X ⊆ V , we have ϕ(X) = 1 if and only if X is a maximal stable set of G,
it is not difficult to see that the above two definitions are equivalent. Equistable graphs
were studied in a series of papers [1, 9, 25, 27, 29, 31, 32, 36]. However, unlike threshold
graphs, the structure of equistable graphs is not understood and the complexity of the
problem of recognizing equistable graphs is open.

Dually Sperner hypergraphs. Sperner hypergraphs can be equivalently defined as the

hypergraphs such that every two distinct hyperedges e and f satisfy

min{|e \ f |, |f \ e|} > 1 . (1)

This observation motivated Chiarelli and Milanič to call in [13] a hypergraph H dually
Sperner if every two distinct hyperedges e and f satisfy

min{|e \ f |, |f \ e|} 6 1 .

Dually Sperner hypergraphs, in general, are not Sperner. The following result was shown
in [13].

Theorem 2 (Chiarelli-Milanič [13]). Every dually Sperner hypergraph is threshold.

We will give an alternative proof of Theorem 2 in Section 5.

1.2 The main definition

The main notion studied in this paper is given by the following.
3
The intersection of two hypergraphs H1 = (V1 , E1 ) and H2 = (V2 , E2 ) is defined in the natural way,
namely as the hypergraph (V1 ∩ V2 , E1 ∩ E2 ).

the electronic journal of combinatorics 26(3) (2019), #P3.18 4

Definition 3. Given a positive integer k, we say that a hypergraph H is k-Sperner if
every two distinct hyperedges e and f satisfy

1 6 min{|e \ f |, |f \ e|} 6 k .

In particular, H is 1-Sperner if every two distinct hyperedges e and f satisfy

min{|e \ f |, |f \ e|} = 1 ,

or, equivalently, if, for any two distinct hyperedges e and f of H with |e| 6 |f |, we have
|e \ f | = 1.

Denoting by Sk the class of all k-Sperner hypergraphs and by S the class of all Sperner
hypergraphs, it is clear that these families of hypergraphs are related by the following chain
of inclusions [
S1 ⊆ S2 ⊆ . . . ⊆ Sk = S .
k>1

The inclusions follow immediately from the definitions, while the equality follows from (1).
Moreover, since we do not allow multiple hyperedges, every 2-uniform hypergraph (that is,
a graph) is k-Sperner for every k > 2. We should therefore not expect useful decomposition
properties for the classes of k-Sperner hypergraphs for k > 2. We focus in this paper on
the case k = 1 and show that hypergraphs in the corresponding subfamily S1 have a nice
structure. Note that by definition, all hypergraphs with at most one hyperedge (possibly
with no vertices) are 1-Sperner. Note also that a hypergraph is 1-Sperner if and only if it
is both Sperner and dually Sperner.
The concept of 1-Sperner hypergraphs already appeared in some graph theoretical
research. Chiarelli and Milanič [12, 13] made use of dually Sperner hypergraphs to char-
acterize two classes of graphs defined by the following properties: every induced subgraph
has a non-negative linear vertex weight function separating the characteristic vectors of
all total dominating sets [12], resp. connected dominating sets [13], from the characteristic
vectors of all other sets. Due to the close relation between 1-Sperner and dually Sperner
hypergraphs (see Observation 13), all the results from [12, 13] can be equivalently stated
using 1-Sperner hypergraphs. In particular, the results of the extended abstract [12] are
stated using the 1-Sperner property in the full version of the paper [14].

1.3 Our results

Our main result is a decomposition theorem for 1-Sperner hypergraphs. We also derive
several consequences of it.

The decomposition theorem. We define a simple operation on hypergraphs called

gluing and show that it produces (with only one small exception) a new 1-Sperner hy-
pergraph from a given pair of 1-Sperner hypergraphs; for an example illustrated with
incidence matrices, see Fig. 2. Conversely, we show that every 1-Sperner hypergraph with
at least one vertex is the gluing of two smaller 1-Sperner hypergraphs; see Theorem 23.

the electronic journal of combinatorics 26(3) (2019), #P3.18 5

Consequences. We use the decomposition theorem to prove the following properties of
1-Sperner hypergraphs.

a) Every 1-Sperner hypergraph is threshold and has a positive threshold separator; see
Theorem 26. In particular, this gives a new, constructive proof of the fact that every
dually Sperner hypergraph is threshold, obtained first by Chiarelli and Milanič in [13];
see Theorem 2.

b) Every 1-Sperner hypergraph is equilizable; see Theorem 27.

c) The characteristic vectors of the hyperedges of a 1-Sperner hypergraph are linearly

independent over the reals; see Theorem 33. It follows that the number of hyperedges
cannot exceed the number of vertices, giving a sharp upper bound on the size of a
1-Sperner hypergraph in terms of its order; see Theorem 34. We also give a sharp
lower bound on the size of a 1-Sperner hypergraph without universal, isolated, and
twin vertices in terms of its order; see Theorem 35.

d) The number of minimal transversals of a 1-Sperner hypergraph is bounded from above

by a quadratic function of its order and they can be efficiently generated; see Theo-
rem 36.

Our study of 1-Sperner hypergraphs is motivated not only by their nice combinatorial
properties but also by their numerous applications in graph theory. Some of them were
already mentioned above and we obtained several others. To keep the length of this
paper reasonable, we decided to present those results in a separate paper [10]. We briefly
summarize them here.
We use the characterizations of so-called threshold and domishold graphs in terms
of forbidden induced subgraphs due to Chvátal and Hammer [16] and Benzaken and
Hammer [4], respectively, to derive further characterizations of these graph classes in
terms of 1-Spernerness, thresholdness, and 2-asummability properties of several related
hypergraphs, namely their vertex cover, clique, independent set, dominating set, and
closed neighborhood hypergraphs.
Furthermore, we use the decomposition theorem for 1-Sperner hypergraphs (Theo-
rem 23) to derive decomposition theorems for four classes of graphs, namely two classes
of split graphs, a class of bipartite graphs, and a class of cobipartite graphs. These decom-
position theorems are based on certain matrix partitions of the corresponding graphs and
give rise to new classes of graphs of bounded clique-width and new polynomially solvable
cases of variants of domination.

1.4 Interrelations between the considered classes of hypergraphs

In Fig. 1, we show the Hasse diagram of the partial order of the hypergraph classes studied
in this paper, ordered with respect to inclusion.
The fact that every 1-Sperner hypergraph is threshold and equilizable is proved in
Theorems 26 and 27, respectively. The fact that every dually Sperner hypergraph is

the electronic journal of combinatorics 26(3) (2019), #P3.18 6

 Sperner

threshold
2-Sperner =
k-asummable for all k ≥ 2 equilizable

2-uniform = graphs

equistable graphs threshold and equilizable

dually Sperner

2-uniform and threshold

=
threshold graphs

1-Sperner

Figure 1: Inclusion relations between several classes of hypergraphs.

threshold was proved by [13]. The fact that every threshold graph is equistable was proved
by [35]. The fact that every equilizable hypergraph is Sperner was proved in Theorem 1.
The remaining inclusions are trivial.
The following examples show that all inclusions are strict and there are no other
inclusions:

• the complete graph K4 is a 2-uniform hypergraph that is threshold but not dually
Sperner;

• the hypergraph with vertex set {1, 2, 3} and hyperedge set {{1, 2, 3}} is 1-Sperner
but not 2-uniform;

• the hypergraph with vertex set {1} and hyperedge set {∅, {1}} is dually Sperner
but not Sperner,

• an equilizable hypergraph that is not threshold is presented in Example 28;

• a 2-uniform threshold hypergraph that is not equilizable is presented in Example 29;

• a threshold and equilizable hypergraph that is neither dually Sperner nor 2-Sperner
is the complete 3-uniform hypergraph H6,3 ; see Example 30;

the electronic journal of combinatorics 26(3) (2019), #P3.18 7

 • the cycle C4 is an equistable graph that, when viewed as a 2-uniform hypergraph,
is not threshold [16];

• the path P4 is a graph that is not equistable [35]; moreover, it is also a Sperner
hypergraph that is not equilizable.

The remaining non-inclusions follow by transitivity.

Structure of the paper. In Section 2, we collect the necessary definitions and prelimi-
nary results. We also consider several operations on hypergraphs and show that the class
of 1-Sperner hypergraphs is (almost always) closed under these operations. In Section 3
we give a necessary condition for a uniform hypergraph to be 1-Sperner and identify two
families of uniform 1-Sperner hypergraphs. Building on the results of Sections 2 and 3,
we develop in Section 4 the composition theorem for 1-Sperner hypergraphs. Various
consequences of this theorem are examined in Section 5.

2 Definitions and hypergraph operations

The order and the size of a hypergraph H refer to the number of its vertices, resp. hy-
peredges. Every hypergraph H = (V, E) with a fixed pair of orderings of its vertices and
edges, say V = {v1 , . . . , vn }, and E = {e1 , . . . , em }, can be represented with its incidence
matrix AH ∈ {0, 1}E×V having rows and columns indexed by edges and vertices of H,
respectively, and defined as 
H 1, if vj ∈ ei ;
Ai,j =
0, otherwise.
Note the slight abuse of notation above: the incidence matrix does not depend only on
the hypergraph but also on the pair of orderings of its vertices and edges. We will be able
to neglect this technical issue in the paper, often but not always.
k-asummable hypergraphs. A hypergraph is k-asummable if it has no k (not nec-
essarily distinct) independent sets A1 , . . . , Ak and k (not necessarily distinct) dependent
sets B1 , . . . , Bk such that
Xk Xk
χ Ai = χBi .
i=1 i=1

A hypergraph is asummable if it is k-asummable for every k > 2. The following characteri-

zation of threshold graphs follows from analogous characterizations of threshold monotone
Boolean functions; see [18].

Theorem 4 (Chow [15] and Elgot [20]). A hypergraph is threshold if and only if it is
asummable.

Next we consider several operations on hypergraphs and show that the class of 1-
Sperner hypergraphs is (almost always) closed under these operations.

the electronic journal of combinatorics 26(3) (2019), #P3.18 8

2.1 Hypergraph complementation
Given a hypergraph H = (V, E), the complement of H is the hypergraph H with V (H) =
V and E(H) = {e | e ∈ E(H)}, where ē denotes V \ e for any subset e ⊆ V .
Proposition 5. The complement of every 1-Sperner hypergraph is 1-Sperner.
Proof. This follows directly from the definition, using the fact that for every two sets
e, f ⊆ V , we have e \ f = f \ e and f \ e = e \ f .
As the next example shows, the closure under complementation does not hold for the
classes of threshold Sperner hypergraphs and 2-asummable Sperner hypergraphs.
Example 6. Consider the 3-uniform hypergraph H = (V, E) with V = {1, . . . , 6} in
which a set e = {x, y, z} ⊆ V forms a hyperedge if and only if e contains at least two
elements of {1, 2, 3, 4}. Then H is a threshold hypergraph, with a threshold separator
(w, t) given by (w(1), . . . , w(6)) = (3, 3, 3, 3, 1, 1) and t = 7. Since H is threshold, it is
also 2-asummable by Theorem 4. Its complement is the hypergraph H = (V, E) with
E = {e ⊆ V : |e| = 3, e * {1, 2, 3, 4}}. Since in H, sets A1 = {1, 2, 3, 4} and A2 = {5, 6}
are independent, while sets B1 = {1, 2, 5} and B2 = {3, 4, 6} are hyperedges, such that
χA1 + χA2 = χB1 + χB2 , we infer that H is not 2-asummable, hence also not threshold.

2.2 Gluing of hypergraphs

The decomposition theorem (Theorem 23) is based on the following general operation.
Definition 7 (Gluing of two hypergraphs). Given a pair of vertex-disjoint hypergraphs
H1 = (V1 , E1 ) and H2 = (V2 , E2 ) and a new vertex z 6∈ V1 ∪ V2 , the gluing of H1 and H2
is the hypergraph H = H1 H2 such that
V (H) = V1 ∪ V2 ∪ {z}
and
E(H) = {{z} ∪ e | e ∈ E1 } ∪ {V1 ∪ e | e ∈ E2 } .
Let us note the operation is not commutative and is well-defined also if some of
the sets V1 , V2 , E1 , and E2 are empty. The operation can be visualized easily in terms
of incidence matrices. Let ni = |Vi | and mi = |Ei | for i = 1, 2, and let us denote by 0k,` ,
resp. 1k,` , the k × ` matrix of all zeroes, resp. of all ones. Assuming that the order of
vertices in AH1 H2 is (z, V1 , V2 ), the incidence matrix of the gluing of H1 and H2 can be
written as  m ,1 
H1 H 2 1 1 AH1 0m1 ,n2
A = .
0m2 ,1 1m2 ,n1 AH2
See Fig. 2 for an example.
To further illustrate the operation of gluing, let us note that the operation generalizes
the operations of adding an isolated or a universal vertex. A vertex u in a hypergraph
H = (V, E) is said to be universal (resp., isolated ) if it is contained in all (resp., in no)
hyperedges. The operations of adding an isolated or a universal vertex to a hypergraph
are defined in the natural way.

the electronic journal of combinatorics 26(3) (2019), #P3.18 9

  
A H1
=
1 0 0  z 
0 1 0 1 1 0 0 0 0 0 0
 1 0 1 0 0 0 0 0 
   
1 1 1 0 A H1 H2
=
 0 1 1 1 1 1 1 0 

 0 1 1 1 1 1 0 1 
A H2
= 1 1 0 1 
1 0 1 1 0 1 1 1 1 0 1 1

Figure 2: An example of gluing of two hypergraphs.

Observation 8. For every hypergraph H, the following holds:

• The hypergraph obtained by adding an isolated vertex to H is the result of gluing of

(∅, ∅) and H.

• The hypergraph obtained by adding a universal vertex to H is the result of gluing of

H and (∅, ∅).

An immediate consequence of the definitions is that the gluing and complementation

operations are related as follows.

Observation 9. If H = H1 H2 , then H = H2 H1 (assuming that in both gluing

operations the same new vertex is used).

The following observation is also easy to see.

Observation 10. If the gluing of H1 and H2 is a 1-Sperner hypergraph, then H1 and H2

are also 1-Sperner.

The next proposition establishes a partial converse. Gluing preserves 1-Spernerness,

unless the resulting hypergraph is not Sperner, which happens only in one very special
case.

Proposition 11. For every pair H1 = (V1 , E1 ) and H2 = (V2 , E2 ) of vertex-disjoint

1-Sperner hypergraphs, their gluing H1 H2 is a 1-Sperner hypergraph, unless E1 = {V1 }
and E2 = {∅} (in which case the hypergraph H1 H2 is not Sperner).

Proof. Let e and f be two distinct edges of H1 H2 . If z ∈ e ∩ f then their differences

are the same as the corresponding differences of e \ {z} and f \ {z}, both of which are
hyperedges of H1 . If z 6∈ e ∪ f then their differences are the same as the corresponding
differences of e \ V1 and f \ V1 , both of which are hyperedges of H2 . If z ∈ e \ f , then
e \ f = {z} and f \ e 6= ∅, unless e = V1 , and f = ∅ (which implies E1 = {V1 } and
E2 = {∅} by our assumption that both H1 and H2 are 1-Sperner). The case of z ∈ f \ e
is symmetric.

the electronic journal of combinatorics 26(3) (2019), #P3.18 10

2.3 Sperner reductions and hypergraph transversals
Given a hypergraph H = (V, E), its Sperner reduction, Sp(H), is the hypergraph with
vertex set V and with hyperedges the inclusion-minimal elements of E.
The following observation is easy to prove from the definitions.
Observation 12. For any two hypergraphs H1 and H2 , if E(Sp(H1 )) ⊆ E(H2 ) ⊆ E(H1 ),
then E(Sp(H1 )) = E(Sp(H2 )).
The following observation is a direct consequence of the definitions of dually Sperner
and 1-Sperner hypergraphs.
Observation 13. The Sperner reduction of every dually Sperner hypergraph is 1-Sperner.
Furthermore, the problem of studying the thresholdness property in a class of hyper-
graphs reduces to the class of their Sperner reductions.
Proposition 14. Let H = (V, E) be a hypergraph, let w : V → Z>0 and t ∈ Z>0 . Then,
(w, t) is a threshold separator of H if and only if (w, t) is a threshold separator of Sp(H).
In particular, H is threshold if and only if its Sperner reduction is threshold.
Proof. Let us call a subset of vertices X ⊆ V heavy if w(X) > t, and light, otherwise. The
pair (w, t) is a threshold separator of H if and only if the heavy subsets of V are precisely
those containing a hyperedge of H. Since the set of heavy subsets depends only on (w, t)
and not on H and a subset of V contains a hyperedge of H if and only if it contains a
hyperedge of Sp(H), the proposition follows.
Let H = (V, E) be a hypergraph. A transversal of H is a set of vertices intersecting
all hyperedges of H. The transversal hypergraph HT is the hypergraph with vertex set V
in which a set X ⊆ V is a hyperedge if and only if X is an inclusion-minimal transversal
of H. (In particular, if H has no hyperedge, then its transversal hypergraph is HT =
(V (H), {∅}).)
Observation 15 (see, e.g., Berge [6]). If H is a Sperner hypergraph, then (HT )T = H.
A pair of a mutually transversal Sperner hypergraphs naturally corresponds to a pair
of dual monotone Boolean functions, see [18].
The next proposition, which will be used in the proof of Theorem 36, describes how
to compute the transversal hypergraph of the gluing of two hypergraphs H1 and H2 from
their transversal hypergraphs.
Proposition 16. Let H be a gluing of two vertex-disjoint hypergraphs H1 = (V1 , E1 ) and
H2 = (V2 , E2 ) with V (H) = V1 ∪ V2 ∪ {z}. Then,
 


 T T

 Sp E H1 ∪ {z} ∪ e | e ∈ E H2 ∪ {{z, u} | u ∈ V1 } , if E1 6= ∅;
E HT = 


 Sp E HT ∪ {{u} | u ∈ V } , if E = ∅.
2 1 1

the electronic journal of combinatorics 26(3) (2019), #P3.18 11

Proof. Let
(


E H1T ∪ {z} ∪ e | e ∈ E H2T ∪ {{z, u} | u ∈ V1 } , if E1 =
6 ∅;
F =

E H2T ∪ {{u} | u ∈ V1 } , if E1 = ∅.

We will first show that every set in F is a transversal of H and then we will argue that
every minimal transversal of H appears in F . Together, by Theorem 12, these two claims
will imply the stated equality.
The first claim is easy to see by the definition of the gluing operation.
For the second claim, let X be a minimal transversal of H. Suppose first that E1 = ∅.
Note that in this case z is an isolated vertex of H, so no minimal transversal of H can
contain z. If X ∩ V1 6= ∅, then X = {u} for some u ∈ V1 by the minimality property. If
X ∩ V1 = ∅, then X must be a minimal transversal of H.
Finally, assume that E1 6= ∅. Suppose also that V1 = ∅ (and hence E1 = {∅}). Then
all minimal transversals of H must contain z and must intersect
all hyperedges of H2 .
T
Thus, X must have the form X = {z} ∪ e for some e ∈ E H2 ; in particular X ∈ F .
Now let V1 6= ∅. If z 6∈ X, then X must be a minimal transversal of H1 . If z ∈ X and
X ∩ V1 6= ∅, then by minimality we must have X = {z, u} for some u
∈ V1 . If z ∈ X and
X ∩ V1 = ∅, then we must have X = {z} ∪ e for some e ∈ E H2T . In either case, X
belongs to F . This completes the proof.

2.4 Ungluing hypergraphs

We next introduce some terminology related to hypergraphs that are the result of a
gluing operation. Given a vertex z of a hypergraph H, we say that a hypergraph H
is z-decomposable if for every two hyperedges e, f ∈ E(H) such that z ∈ e \ f , we have
e\{z} ⊆ f . Equivalently, if the vertex set of H can be partitioned as V (H) = {z}∪V1 ∪V2
such that H = H1 H2 for some hypergraphs H1 = (V1 , E1 ) and H2 = (V2 , E2 ). We call
H = H1 H2 a z-decomposition of H.
The following proposition gathers some basic properties of decomposability.

Proposition 17. Let H be a hypergraph. Then, the following holds:

(i) If z is a vertex of H such that H is z-decomposable, then H is also z-decomposable.

(ii) If z is an isolated or a universal vertex of H, then H is z-decomposable.

Proof. Statement (i) follows from Observation 9.

Statement (ii) is related to Observation 8. Note that z is universal in H if and only
if it is isolated in H. By (i), it therefore suffices to prove the statement for the case
when z is an isolated vertex of H. In this case, the column of AH indexed by z is the all
zero vector. It follows that H is z-decomposable, as follows: V (H) = {z} ∪ V1 ∪ V2 with
H = H1 H2 , H1 = (V1 , E1 ) and H2 = (V2 , E2 ), where V1 = E1 = ∅, V2 = V \ {z}, and
E2 = E(H).

the electronic journal of combinatorics 26(3) (2019), #P3.18 12

 Recall that by Theorem 10, if a 1-Sperner hypergraph H has a z-decomposition H =
H1 H2 , then H1 and H2 are also 1-Sperner.
Whether a given hypergraph is z-decomposable for some vertex z can be checked in a
straightforward way in polynomial time.
Proposition 18. Let H = (V, E) be a hypergraph with V 6= ∅ and E 6= ∅ given by the
lists of its vertices and hyperedges. We can recognize if H is z-decomposable for some
z ∈ V and find a corresponding z-decomposition H = H1 H2 (if there is one) in time
O(|V |2 |E|).
Proof. It suffices to show that for a given vertex z ∈ V we can verify in time O(|V ||E|)
if H is z-decomposable and find a corresponding z-decomposition H = H1 H2 (if there
is one).
First, we partition the hyperedges of H into those containing z and those not containing
z. Secondly, we compute the sets E1 = {e \ {z} | z ∈ e ∈ E} and V1 = ∪{e | e ∈ E1 }.
Thirdly, we verify if for every hyperedge e ∈ E not containing z, we have V1 ⊆ e. If this
condition is not satisfied, then H is not z-decomposable. If the condition is satisfied, then
we compute the sets V2 = V \ (V1 ∪ {z}) and E2 = {e \ V1 | z 6∈ e ∈ E}. We return the
z-decomposition H = H1 H2 , where P H1 = (V1 , E1 ) and H2 = (V2 , E2 ). Since each of the
steps can be performed in time O( e∈E |e|) = O(|V ||E|), the claimed time complexity
follows.
Clearly, if H = (V, E) is a hypergraph with at least one vertex and no hyperedges,
then H is z-decomposable for every z ∈ V and a z-decomposition of H can be computed
in time O(|V |).

3 Uniform 1-Sperner hypergraphs

In the next lemma we give a necessary condition for a uniform hypergraph to be 1-Sperner.
The condition will be used in the proof of Theorem 23.
Lemma 19. Let H be a k-uniform 1-Sperner hypergraph, where k > 1. Then, either
there is a subset P of vertices of size k − 1 such that P ⊆ e for all e ∈ E(H) or there is
a subset Q of vertices of size k + 1 such that e ⊆ Q for all e ∈ E(H).
Proof. The statement of the lemma holds if H has at most one hyperedge. So let us
assume that H has at least two hyperedges, say e and f . Let P = e ∩ f . Since H is
1-Sperner, |P | = k − 1. If all hyperedges of H contain P , then we are done.
If there is a hyperedge g such that P * g, say u ∈ P \ g, then e and f are the only
hyperedges containing P , since otherwise g would contain all vertices of such hyperedges
other than u, which would imply |g| > k. Consequently, all hyperedges that miss a vertex
of P are subsets of Q = e ∪ f , and the lemma is proved.
Lemma 19 suggests the following two families of uniform 1-Sperner hypergraphs (with
P = X and Q = X ∪ Y ).

the electronic journal of combinatorics 26(3) (2019), #P3.18 13

Example 20. Given k > 1, a k-star is a k-uniform hypergraph H = (V, E) such that
there exists sets X, Y ⊆ V such that

• X ∪ Y ⊆ V where |X| = k − 1, Y 6= ∅, and X ∩ Y = ∅, and

• E = {X ∪ {y} | y ∈ Y }.

If this is the case, we say that H is the (k-)star generated by (V, X, Y ).

Clearly, every k-star is 1-Sperner. Moreover, let us verify that every k-star is z-
decomposable with respect to every vertex z. Let H be a k-star generated by (V, X, Y )
and let z ∈ V (H). If z ∈ X, then k > 2 and we have H = H1 H2 where H1 is the
(k − 1)-star generated by X \ {z} and Y and V (H2 ) = E(H2 ) = ∅. If z ∈ Y , then we have
H = H1 H2 where H1 = (X, {X}) and H2 = (Y \ {z}, {{y} | y ∈ Y \ {z}}). Finally, if
z ∈ V \ (X ∪ Y ), then z is isolated and H is z-decomposable by Proposition 17.

Example 21. Given k > 1, a k-antistar is a k-uniform hypergraph H = (V, E) such that
there exists sets X, Y ⊆ V such that

• X ∪ Y ⊆ V , Y 6= ∅, X ∩ Y = ∅, and |X ∪ Y | = k + 1, and

• E = {X ∪ (Y \ {y}) | y ∈ Y }.

If this is the case, we say that H is the (k-)antistar generated by (V, X, Y ). Note that
every antistar is the complement of a star. It follows, using Propositions 5 and 17 and
the properties of stars observed in Example 20, that every antistar is 1-Sperner and z-
decomposable with respect to each vertex z.

4 Decomposition theorem
To prove the main structural result about 1-Sperner hypergraph (Theorem 23), we need
the following technical lemma.

Lemma 22. Let H be a 1-Sperner hypergraph with E(H) 6= ∅ and let C be a hyperedge
of H of maximum size. Then, for every two distinct vertices x, y 6∈ C and every two
hyperedges A containing x and B containing y, |A| 6 |B| implies A ∩ C ⊆ B ∩ C.

Proof. Note that |A| 6 |C|, therefore A \ C = {x}, since H is 1-Sperner. Analogously,
B \ C = {y}. Thus, if the sets A ∩ C and B ∩ C were not comparable with respect to
inclusion, the pair {A, B} would violate the 1-Sperner property of H.

Theorem 23. Every 1-Sperner hypergraph H = (V, E) with V 6= ∅ is z-decomposable for

some z ∈ V (H), that is, it is the gluing of two 1-Sperner hypergraphs.

Proof. By Proposition 17, we may assume that H does not have any isolated vertices.
For every v ∈ V , let
k(v) = max |e|
v∈e∈E

the electronic journal of combinatorics 26(3) (2019), #P3.18 14

and let k(H) = maxv∈V k(v).
We consider two cases.
Case 1: Not all the k(v) values are the same. Let v ∈ V be a vertex with the smallest
k(v) value. Then k(v) < k(H) by the assumption of this case.
First we show that for every hyperedge f ∈ E such that v 6∈ f , we have |f | > k(v).
Suppose for a contradiction that there exists a hyperedge f ∈ E such that v 6∈ f and
|f | < k(v). Let e be a hyperedge containing v of size k(v), and let g be a hyperedge of
maximum size, that is, |g| = k(H). Then v 6∈ g, since k(v) < |g|. Since H is Sperner, there
exists a vertex u ∈ f \ g. Note that |f | 6 |g|, therefore f \ g = {u}, since H is 1-Sperner.
Moreover, u 6= v since u ∈ f and f does not contain v. We know that k(u) > k(v), by
our choice of v. Furthermore, there exists a hyperedge h containing u and of size k(u).
Since |h| = k(u) > k(v) > |f |, we have h 6= f . Applying Lemma 22 with (x, y, A, B, C) =
(u, v, f, e, g) yields f ∩ g ⊆ e ∩ g. Applying Lemma 22 with (x, y, A, B, C) = (v, u, e, h, g)
yields e ∩ g ⊆ h ∩ g. Consequently, f ∩ g ⊆ h ∩ g. On the other hand, f \ g = h \ g = {u}.
It follows that f ⊆ h, contradicting the Sperner property of H.
To complete Case 1, we show that H is v-decomposable. This is because for every two
hyperedges e, f ∈ E such that v ∈ e \ f , we have |f | > k(v) > |e|, implying |f \ e| > |e \ f |,
from what we derive, using the fact that H is 1-Sperner, that |e\f | = 1, that is, e\{v} ⊆ f .
Case 2: All the k(v) values are the same. Let k = k(H). If k 6 1, then H is z-
decomposable with respect to every vertex z. So suppose that k > 2. Consider the
subhypergraph H0 of H with V (H0 ) = V (H) formed by the hyperedges of H of size k.
By Lemma 19 applied to H0 , either there is a subset P of vertices of size k − 1 such that
P ⊆ e for all e ∈ E(H0 ) or there is a subset Q of vertices of size k + 1 such that e ⊆ Q
for all e ∈ E(H0 ).
Suppose first that there is a subset P of vertices of size k − 1 such that P ⊆ e for all
e ∈ E(H0 ). If H0 = H, that is, all hyperedges of H are of size k, then H is z-decomposable
with respect to every vertex z (cf. Example 20). So we may assume that H0 6= H, that is,
that H contains a hyperedge g of size less than k.* By the assumption of Case 2, we know
that g ⊆ ∪f ∈E(H0 ) f . Since H is Sperner, g is not contained in any of the
hyperedges of H0 ;
moreover g contains at least two vertices from the set Y = ∪f ∈E(H0 ) f \ P . If g contains
at least three vertices from Y , say y1 , y2 , y3 , then the hyperedges P ∪ {y1 } and g would
violate the 1-Sperner property, since {y2 , y3 } ⊆ g \ (P ∪ {y1 }) and P \ g ⊆ (P ∪ {y1 }) \ g
(note that |P \ g| > 3). It follows that |g ∩ Y | = 2. In fact, we have |Y | = 2, say
Y = {y1 , y2 }, since otherwise, using similar arguments as above, we see that the sets
P ∪ {y} and g would violate the 1-Sperner property, where y ∈ Y \ g. It follows that H0
has exactly 2 hyperedges, and Y ⊆ e for every set e ∈ E(H) \ E(H0 ). Consequently, H
is y-decomposable for every y ∈ Y . Decomposing H with respect to y = y1 , for instance,
we have H = H1 H2 where H1 = (V1 , E1 ) with V1 = V \ {y1 }, E1 = {e \ {y1 } | y1 ∈ e ∈
E(H)}, and H2 = (∅, {∅}).
It remains to consider the case when there is a subset Q of vertices of size k + 1 such
that e ⊆ Q for all e ∈ E(H0 ). Since we assume that k(v) = k for all vertices v, we
have V = Q. Let us define X = ∩f ∈E(H0 ) f and Y = V \ X. Then, Y 6= ∅, and every
hyperedge g ∈ E(H) \ E(H0 ) must contain Y , since H is Sperner and for every vertex

the electronic journal of combinatorics 26(3) (2019), #P3.18 15

y ∈ Y , the set V \ {y} is a hyperedge of H. Consequently, H is y-decomposable for every
y ∈ Y : taking any y ∈ Y , we have H = H1 H2 where H1 = (V1 , E1 ) with V1 = V \ {y},
E1 = {e \ {y} | y ∈ e ∈ E(H)}, and H2 = (∅, {∅}).
Let us say that a gluing of two vertex-disjoint 1-Sperner hypergraphs H1 = (V1 , E1 ),
H2 = (V2 , E2 ) is safe if it results in a 1-Sperner hypergraph. By Proposition 11, this is
always the case unless E1 = {V1 } and E2 = {∅}. Thus, Theorem 23 and Proposition 11
imply the following composition result for the class of 1-Sperner hypergraphs.

Theorem 24. A hypergraph H is 1-Sperner if and only if it either has no vertices (that
is, H ∈ {(∅, ∅), (∅, {∅})}) or it is a safe gluing of two smaller 1-Sperner hypergraphs.

5 Applications of the decomposition theorem

In this section we present several applications of Theorems 23 and 24 giving further insight
on the properties of 1-Sperner hypergraphs.

5.1 1-Sperner hypergraphs are threshold

Our first application is motivated by the result of Chiarelli and Milanič [13] stating that
every dually Sperner hypergraph is threshold, see Theorem 2. The proof of Theorem 2
given in [13] is based on the characterization of thresholdness in terms of asummability
(Theorem 4) and does not show how to compute a threshold separator of a dually Sperner
hypergraph. Here we give an alternative proof of Theorem 2, based on the composition
theorem of 1-Sperner hypergraphs. In contrast with the proof from [13], our proof is
constructive in the sense that it computes an explicit threshold separator of a 1-Sperner
or, more generally, dually Sperner hypergraph.
Clearly, every 1-Sperner hypergraph is dually Sperner. Therefore, Theorem 2 implies
that every 1-Sperner hypergraph is threshold. We will now derive this fact directly from
the composition theorem. In fact, we will show a bit more, namely that every 1-Sperner
hypergraph H = (V, E) admits a positive threshold separator, that is, a threshold sep-
arator (w, t) such that w : V → Z>0 is a (strictly) positive integer weight function.
The following simple technical claim will be used in the proof.

Lemma 25. For every threshold separator (w, t) of a Sperner threshold hypergraph H =
(V, E), we have:

(i) if w(V ) = t then E = {V }, and

(ii) if t = 0 then E = {∅}.

Proof. If w(V ) = t, then V is a hyperedge and if t = 0, then the empty set is a hyperedge.
(Both of these claims follow from the fact that (w, t) is a threshold separator of H.) In
both cases no other hyperedge may exist due to the Sperner property.

Theorem 26. Every 1-Sperner hypergraph is threshold with a positive threshold separator.

the electronic journal of combinatorics 26(3) (2019), #P3.18 16

Proof. Let H = (V, E) be a 1-Sperner hypergraph. The proof is by induction on n = |V |.
For n = 0, we can obtain a positive threshold separator by taking the (empty) mapping
given by w(x) = 1 for all x ∈ V and the threshold

1, if E = ∅;
t=
0, if E = {∅}.

Now, let n > 1. By Theorem 24, H is the safe gluing of two 1-Sperner hypergraphs,
say H = H1 H2 with H1 = (V1 , E1 ) and H2 = (V2 , E2 ), where V = V1 ∪ V2 ∪ {z},
V1 ∩ V2 = ∅, and z 6∈ V1 ∪ V2 . By the inductive hypothesis, H1 and H2 admit positive
threshold separators. That is, there exist positive integer weight functions wi : Vi → Z>0
and non-negative integer thresholds ti ∈ Z>0 for i = 1, 2 such that for every subset X ⊆ Vi ,
we have wi (X) > ti if and only if e ⊆ X for some e ∈ Ei .
Let us define the threshold t = M w1 (V1 ) + t2 , where M = w2 (V2 ) + 1, and the weight
function w : V → Z>0 by the rule

 M w1 (x), if x ∈ V1 ;
w(x) = w2 (x), if x ∈ V2 ;

M (w1 (V1 ) − t1 ) + t2 , if x = z.

We claim that (w, t) is a positive threshold separator of H. Let us first verify that the so
defined weight function is indeed positive. Since wi for i ∈ {1, 2} are positive and M > 0,
we have w(x) > 0 for all x ∈ V1 ∪ V2 . Moreover, since w1 (V1 ) > t1 , M > 0, and t2 > 0, we
have w(z) > 0. If w(z) = 0, then w1 (V1 ) = t1 and t2 = 0, which by Theorem 25 implies
E1 = {V1 } and E2 = {∅}, contrary to the fact that the gluing is safe. It follows that
w(z) > 0, as claimed.
Next, we verify that (w, t) is a threshold separator of H, that is, that for every subset
X ⊆ V , we have w(X) > t if and only if e ⊆ X for some e ∈ E.
Suppose first that w(X) > t for some X ⊆ V . Let Xi = X ∩ Vi for i = 1, 2. For later
use, we note that
w2 (X2 ) 6 w2 (V2 ) < M . (2)
Suppose first that z ∈ X. Then

M w1 (V1 ) + t2 = t 6 w(X)
= w(z) + w(X1 ) + w(X2 )
= M (w1 (V1 ) − t1 ) + t2 + M w1 (X1 ) + w2 (X2 ) ,

which, using also (2), implies M w1 (X1 ) + w2 (X2 ) > M t1 > M (t1 − 1) + w2 (X2 ) and hence
w1 (X1 ) > t1 . Consequently, there exists e1 ∈ E1 such that e1 ⊆ X1 and the hyperedge
e := {z} ∪ e1 ∈ E satisfies e ⊆ X.
Now, suppose that z 6∈ X. In this case, M w1 (V1 ) + t2 = t 6 w(X) = w(X1 ) + w(X2 ) =
M w1 (X1 ) + w2 (X2 ), which implies

M w1 (X1 ) + w2 (X2 ) > M w1 (V1 ) + t2 . (3)

the electronic journal of combinatorics 26(3) (2019), #P3.18 17

We must have X1 = V1 since if there exists a vertex v ∈ V1 \ X1 , then we would have
M w1 (X1 ) + w2 (X2 ) 6 M w1 (V1 ) − M w1 (v) + w2 (X2 )
6 M w1 (V1 ) − M + w2 (X2 )
< M w1 (V1 ) + t2 ,
where the last inequality follows from (2) and t2 > 0. Therefore, inequality (3) simplifies
to w2 (X2 ) > t2 , and consequently there exists a hyperedge e2 ∈ E2 such that e2 ⊆ X2 .
This implies that H has a hyperedge e := V1 ∪ e2 such that e ⊆ V1 ∪ X2 = X.
For the converse direction, suppose that X is a subset of V such that e ⊆ X for some
e ∈ E. We need to show that w(X) > t. We consider two cases depending on whether
z ∈ e or not. Suppose first that z ∈ e. Then e = {z} ∪ e1 for some e1 ∈ E1 . Due to the
property of w1 , we have w1 (e1 ) > t1 . Consequently,
w(X) > w(e)
= w(z) + w(e1 )
= M (w1 (V1 ) − t1 ) + t2 + M w1 (e1 )
> M w1 (V1 ) − M t1 + t2 + M t1
= M w1 (V1 ) + t2 = t .
Suppose now that z 6∈ e. Then e = V1 ∪ e2 for some e2 ∈ E2 . Due to the property of
w2 , we have w2 (e2 ) > t2 . It follows that w(X) = w(V1 ) + w(e2 ) = M w1 (V1 ) + w2 (e2 ) >
M w1 (V1 ) + t2 = t. This completes the proof.
We now give an alternative proof of Theorem 2 announced above.
An alternative proof of Theorem 2. Let H = (V, E) be a dually Sperner hypergraph.
By Observation 13, its Sperner reduction is 1-Sperner. By Theorem 26, Sp(H) has a
positive threshold separator, say (w, t). Since (w, t) is a threshold separator of Sp(H), it
is also a threshold separator of H, by Theorem 14. Thus, H is threshold.
We would like to emphasize that the above proof implies the following simple efficient
procedure of obtaining a threshold separator of a given dually Sperner hypergraph H:
(1) compute its Sperner reduction, Sp(H), and
(2) construct a positive threshold separator (w, t) of Sp(H) recursively along a decompo-
sition of Sp(H) into smaller 1-Sperner hypergraphs given by Theorem 24 (eventually
resulting in trivial 1-Sperner hypergraphs).
Then (w, t) is a threshold separator of H.

5.2 Further relations between threshold, equilizable, and 1-Sperner hyper-

graphs
The same inductive construction of a threshold separator as that given in the proof of
Theorem 26 shows that every 1-Sperner hypergraph is also equilizable (see Section 1.1 for
the definition).

the electronic journal of combinatorics 26(3) (2019), #P3.18 18

Theorem 27. Every 1-Sperner hypergraph is equilizable.

Theorem 27 can be proved by slightly modifying the above proof of Theorem 26; for
the sake of completeness, we include the proof in Appendix. Theorem 27 will be used in
Section 5.3 to establish an upper bound on the size of a 1-Sperner hypergraph of a given
order.
Combining Theorems 26 and 27 shows that every 1-Sperner hypergraph is threshold
and equilizable. In particular, the properties of thresholdness and equilizability trivially
coincide within the class of 1-Sperner hypergraphs. This raises the question of whether
the two properties are comparable in the larger class of Sperner hypergraphs. As the
following two examples show, this is not the case.

Example 28. The following Sperner hypergraph is equilizable but not threshold: H1 =
(V1 , E1 ) where V1 = {v1 , v2 , v3 , v4 , v5 }, E1 = {{v1 , v2 }, {v2 , v3 , v4 }, {v4 , v5 }}. The function
w : V1 → Z>0 defined by w(v1 ) = 5, w(v2 ) = 4, w(v3 ) = 3, w(v4 ) = 2, and w(v5 ) = 7
assigns a total weight of 9 to each hyperedge and to no other subset of V1 . Thus, H1 is
equilizable. To see that H1 is not threshold, note that any threshold separator (w0 , t0 )
of H1 would have to satisfy w0 (v1 ) + w0 (v2 ) > t0 and w0 (v3 ) + w0 (v4 ) > t0 , as well as
w0 (v1 ) + w0 (v3 ) < t0 and w0 (v2 ) + w0 (v4 ) < t0 , which is impossible. In other words, H1 fails
to be threshold since it is not 2-asummable; cf. Theorem 4.

Example 29. The following Sperner hypergraph is threshold but not equilizable: H2 =
(V2 , E2 ) where V2 = {v1 , v2 , v3 , v4 }, E2 = {{v1 , v2 }, {v1 , v3 }, {v2 , v3 }, {v2 , v4 }, {v3 , v4 }}.
The function w : V2 → Z>0 defined by w(v1 ) = w(v4 ) = 1, w(v2 ) = w(v3 ) = 2, and
threshold t = 3 form a threshold separator of H2 . Thus, H2 is threshold. To see that H2
is not equilizable, note that any function w0 : V2 → Z>0 such that the total weight of every
hyperedge is the same, say t0 , must assign weight t0 /2 to every vertex. Consequently, the
set {v1 , v4 }, which is not a hyperedge, would also be of total weight t0 .

Furthermore, the following examples show that there exist Sperner hypergraphs that
are threshold and equilizable but not 1-Sperner.

Example 30. For every k > 2 and n > 2k, the complete k-uniform hypergraph Hn,k
defined with V (Hn,k ) = {1, . . . , n} and E(Hn,k ) = {X | X ⊆ {1, . . . , n}, |X| = k} is not
1-Sperner, but it is both threshold and equilizable, as verified by the weight function that
is constantly equal 1 and threshold t = k.

5.3 Bounds on the size of 1-Sperner hypergraphs

We now establish some upper and lower bounds on the number of hyperedges in a 1-
Sperner hypergraph with a given number of vertices. By 0, resp. 1, we will denote the
vector of all zeroes, resp. ones, of appropriate dimension (which will be clear from the
context). The following lemma can be easily derived from Theorem 27.

the electronic journal of combinatorics 26(3) (2019), #P3.18 19

Lemma 31. For every 1-Sperner hypergraph H = (V, E) such that E 6= ∅ and E 6= {∅},
there exists a vector x ∈ RV>0 such that AH x = 1 and 1> x > 1.

Proof. Let H = (V, E) be a 1-Sperner hypergraph as in the statement of the lemma. By

Theorem 27, H is equilizable. Let w : V → Z>0 be a non-negative integer weight function
and t ∈ Z>0 a non-negative integer threshold such that for every subset X ⊆ V , we have
w(X) = t if and only if X ∈ E. If t = 0, then ∅ ∈ E and consequently E = {∅},
a contradiction. It follows that t > 0, and we can define the vector x ∈ RV>0 given by
xv = w(v)/t for all v ∈ V . We claim that vector x satisfies the desired properties AH x = 1
and 1> x > 1.
Since w(X) = t for all X ∈ E, we have AH x = 1. Since E 6= ∅, an arbi-
trary hyperedge
P e ∈ E shows that t = w(e) 6 w(V ). Consequently, we also have
>
1 x = v∈V xv = w(V )/t > 1 .

Corollary 32. For every 1-Sperner hypergraph H = (V, E) and every vector λ ∈ RE we
have
λ> AH = 1> ⇒ λ> 1 > 1 .

Proof. If E = ∅ or E = {∅}, then the left hand side of the above implication is always
false. In all other cases, by Lemma 31, there exists a vector x ∈ RV such that AH x = 1
and 1> x > 1. Therefore, equation λ> AH = 1> implies λ> 1 = λ> AH x = 1> x > 1.
The composition theorem and the above corollary imply the following useful property
of 1-Sperner hypergraphs.

Theorem 33. For every 1-Sperner hypergraph H = (V, E) such that E 6= {∅}, the char-
acteristic vectors of its hyperedges are linearly independent (over the field of real numbers).

Proof. We use induction on |V |. If |V | 6 1, then the statement holds since E 6= {∅}.

Suppose now that |V | > 1. Then by Theorem 23, H is the gluing of two 1-Sperner
hypergraphs, say H = H1 H2 with H1 = (V1 , E1 ) and H2 = (V2 , E2 ), where V =
V1 ∪ V2 ∪ {z}, V1 ∩ V2 = ∅, and z 6∈ V1 ∪ V2 .
Let λ ∈ RE be a vector such that λ> AH = 0. Let λ1 and λ2 be the restrictions of
λ to the hyperedges corresponding to E1 and E2 , respectively. The equation λ> AH = 0
implies the system of equations

(λ1 )> 1 = 0 ∈ R ,
(λ1 )> AH1 + ((λ2 )> 1)1> = 0> ∈ RV1 ,
(λ2 )> AH2 = 0> ∈ RV2 .

In all cases, the inductive hypothesis implies that λ1 = 0> ∈ RE1 and λ2 = 0> ∈ RE2 ,
except in the case when E2 = {∅}. In this case, λ2 is a single number, say λ∗ . If λ∗ = 0,
then λ1 = 0> follows by the induction hypothesis. If λ∗ 6= 0, then λ̂ := −λ1 /λ∗ satisfies
λ̂> AH1 = 1> and λ̂> 1 = 0, contradicting Corollary 32.

the electronic journal of combinatorics 26(3) (2019), #P3.18 20

 Since the characteristic vectors of the hyperedges of an n-vertex 1-Sperner hypergraph
are linearly independent vectors in Rn , we obtain the following upper bound on the size
of a 1-Sperner hypergraph in terms of its order.

Corollary 34. For every 1-Sperner hypergraph H = (V, E) with V 6= ∅, we have |E| 6
|V |.

The bound |E| 6 |V | can also be proved more directly from the decomposition theorem
(Theorem 24), using induction on the number of vertices and analyzing various cases
according to whether the two constituent hypergraphs have non-empty vertex set or not.
We decided to include the proof based on Theorem 33, since linear independence is an
interesting property of 1-Sperner hypergraphs and the inequality |E| 6 |V | is just one
consequence of that.
We now turn to the lower bound. Recall that a vertex u in a hypergraph H = (V, E)
is said to be universal (resp., isolated ) if it is contained in all (resp., in no) hyperedges.
Moreover, two vertices u, v of a hypergraph H = (V, E) are twins if they are contained in
exactly the same hyperedges.
Corollary 34 gives an upper bound on the size of a 1-Sperner hypergraph in terms of
its order. Can we prove a lower bound of a similar form? In general not, since adding
universal vertices, isolated vertices, or twin vertices preserves the 1-Sperner property and
the size, while it increases the order. However, as we show next, for 1-Sperner hypergraphs
without universal, isolated, and twin vertices, the following sharp lower bound on the size
in terms of the order holds.

Proposition 35. For every 1-Sperner hypergraph H = (V, E) with |V | > 2 and without
universal, isolated, and twin vertices, we have the following sharp lower bound
 
|V |
|E| > + 1.
2

Proof. We use induction on n = |V |. For n ∈ {2, 3, 4}, it can be easily verified that the
statement holds.
Now, let H = (V, E) be a 1-Sperner hypergraph with n > 5 and without universal
vertices, isolated vertices, and twin vertices. By Theorem 23, H is the gluing of two
1-Sperner hypergraphs, say H = H1 H2 with H1 = (V1 , E1 ) and H2 = (V2 , E2 ), where
V = V1 ∪ V2 ∪ {z}, V1 ∩ V2 = ∅, and z 6∈ V1 ∪ V2 . Since H has no twins, H1 and H2 also
have no twins. Let ni = |Vi | and mi = |Ei | for i = 1, 2, and let m = |E|.
We have m = m1 +m2 , and by the rules of the gluing, n = n1 +n2 +1. By Proposition 5,
we may assume that n1 > n2 (otherwise, we can consider the complementary hypergraph).
In particular, n1 > 3. The fact that H does not have a universal vertex implies H1 does
not have a universal vertex. Similarly, H2 does not have an isolated vertex. Since H does
not have any pairs of twin vertices, we have that either H1 does not have a isolated vertex,
or H2 does not have a universal vertex. We may assume that H2 does not have a universal
vertex (otherwise, we consider a different gluing in which we delete the universal vertex

the electronic journal of combinatorics 26(3) (2019), #P3.18 21

from H2 and add an isolated vertex to H1 ). Since H2 is a Sperner hypergraph without
an isolated or a universal vertex, we have n2 6= 1.
Suppose first that n2 > 2. We apply the inductive hypothesis for H10 and H2 , where H10
is the hypergraph obtained from H1 by deleting from it the isolated vertex (if it exists).
Letting n01 = |V (H10 )| and m01 = |E(H10 )|, we thus have n01 > n1 − 1 and also n01 > 2. We
obtain
n0 + 2 n1 + 1
m1 = m01 > 1 >
2 2
and
n2
m2 > + 1.
2
Consequently,
n1 + 1 n2 + 2 n1 + n2 + 3 n
m = m1 + m2 > + = = + 1,
2 2 2 2
and, since m is integer, the desired inequality
lnm
m> +1
2
follows.
Suppose now that n2 = 0. In this case, since H does not have a universal vertex,
we must have E2 = {∅} and m2 = 1. As above, let H10 be the hypergraph obtained
from H1 by deleting from it the isolated vertex (if it exists). Letting n01 = |V (H10 )| and
m01 = |E(H10 )|, we obtain, by applying the inductive hypothesis to H10 ,

n01 n1 + 1
m1 = m01 > +1> ,
2 2
which implies
n1 + 1 n n
m = m1 + 1 > + 1 = + 1 = + 1.
2 2 2
This completes the proof of the inequality.
To see that the inequality is sharp, consider the following recursively defined family
of hypergraphs Hk for k > 2:

• H2 = ({v1 , v2 }, {{v1 }, {v2 }}).

0 0
• For k > 2, we set Hk = Hk−1 Hk−1 where Hk−1 is the hypergraph obtained from
a disjoint copy of Hk−1 by adding to it an isolated vertex.

An inductive argument shows that for every k > 2, we have nk = |V (Hk )| = 2k − 2,

mk = |E(Hk )| = 2k−1 , and consequently mk = n2k + 1 .

the electronic journal of combinatorics 26(3) (2019), #P3.18 22

5.4 Minimal transversals of 1-Sperner hypergraphs
Recall that a transversal of H is a set of vertices intersecting all hyperedges of H.
Theorem 36. The number of minimal transversals of every 1-Sperner hypergraph H =
(V, E) is at most   
|V |
max 1, |V |, .
2
This bound is sharp. Moreover, the family of minimal transversals of a given 1-Sperner
hypergraph H = (V, E) can be generated in time O(|V |3 |E|).
Proof. We first prove the upper bound on the size of the transversal hypergraph HT . We
use induction on n = |V |. The claim is clear for n = 0. For n ∈ {1, 2, 3}, the claim is
that every 1-Sperner hypergraph of order n has at most n minimal transversals. This is
true since for these small values of n, no family of pairwise incomparable subsets of an
n-element set can have more than n elements.
Now, let H = (V, E) be a 1-Sperner hypergraph with n > 4. By Theorem 23, H
is the gluing of two 1-Sperner hypergraphs, say H = H1 H2 with H1 = (V1 , E1 ) and
H2 = (V2 , E2 ), where V = V1 ∪ V2 ∪ {z}, V1 ∩ V2 = ∅, and z 6∈ V1 ∪ V2 . Denoting

ni = |Vi | for i ∈ {1, 2}, the inductive hypothesis implies that |E(HiT )| 6 max{1, ni , n2i }
for i ∈ {1, 2}. Suppose first that n2 = 0. In this case, E(H2T ) is either ∅ (if E(H2 ) = {∅})
or {∅} (if E(H2 ) = ∅). By Theorem 16, it suffices to show the inequality
     
n1 n
max 1, n1 , + max {1, n1 } 6 max 1, n, .
2 2



Using n > 4 and consequently n1 > 3, the inequality reduces to n21 + n1 6 n12+1 , which
is satisfied with equality.
Suppose now that n2 > 1. By Theorem 16, it suffices to show the inequality
        
n1 n2 n
max 1, n1 , + max 1, n2 , + n1 6 max 1, n, .
2 2 2
The inequality holds for any n > 4 and any n1 , n2 such that n2 > 1 and n1 + n2 + 1 = n.
The details are left to the reader.
To see that the inequality is sharp, let n > 3 and consider the family Hn,n−1 of complete
(n − 1)-uniform hypergraphs; see Example 30. It is clear that the hypergraph Hn,n−1 is

transversal hypergraph is the complete 2-uniform hypergraph Hn,2 , which
1-Sperner. Its
is of size n2 .
It remains to show that the transversal hypergraph HT of a given 1-Sperner hypergraph
H = (V, E) with n = |V | and m = |E| can be generated in time O(n3 m). We may assume
that n > 1. The algorithm is as follows:
1. First, we compute the z-decomposition H = H1 H2 (for some z ∈ V ).
2. Secondly, for i ∈ {1, 2}, we recursively generate the transversal hypergraphs H1T
and H2T .

the electronic journal of combinatorics 26(3) (2019), #P3.18 23

 3. Finally, we compute the transversal hypergraph HT using Theorem 16.
Let T (n, m) denote the running time of this algorithm. Step 1 of the algorithm can be
done in time O(n2 m) by Theorem 18. For Step 2, let H1 = (V1 , E1 ) and H2 = (V2 , E2 ),
where V = V1 ∪ V2 ∪ {z}, V1 ∩ V2 = ∅, and z 6∈ V1 ∪ V2 . Denoting ni = |Vi | and mi = |Ei |
and for i ∈ {1, 2}, we can do Step 2 in time T (n1 , m1 ) + T (n2 , m2 ). For Step 3, let F
be defined as in the proof of Theorem 16, and notice that F is not equal to its Sperner
reduction if and only if one of the following happens: (i) the empty set is a minimal
transversal of H1 , (ii) the empty set is a minimal transversal of H2 , or (iii) {u} is a
minimal transversal of H1 for some u ∈ V1 . These conditions can be verified either in
constant time (in cases (i) and (ii)) or in O(|E(H1T )|) = O(n21 ) time (in case (iii)).
The above reasoning leads to the inequality T (n, m) 6 O(n2 m)+T (n1 , m1 )+T (n2 , m2 ).
Since n = n1 + n2 + 1 and m = m1 + m2 , this inequality implies that T (n, m) = O(n3 m),
as claimed.

Acknowledgements
The authors are grateful to the anonymous reviewers for their comments and to Nina
Chiarelli and Sylwia Cichacz for helpful discussions. The work for this paper was done
in the framework of bilateral projects between Slovenia and the USA, partially financed
by the Slovenian Research Agency (BI-US/14–15–050, BI-US/16–17–030, and BI-US/18–
19–029).

References
[1] Liliana Alcón, Marisa Gutierrez, István Kovács, Martin Milanič, and Romeo Rizzi.
Strong cliques and equistability of EPT graphs. Discrete Appl. Math., 203:13–25,
2016. doi:10.1016/j.dam.2015.09.016.
[2] J. M. Anthonisse. A note on equivalent systems of linear Diophantine equations. Z.
Operations Res. Ser. A-B, 17:A167–A177, 1973.
[3] Amos Beimel and Enav Weinreb. Monotone circuits for monotone
weighted threshold functions. Inform. Process. Lett., 97(1):12–18, 2006.
doi:10.1016/j.ipl.2005.09.008.
[4] C. Benzaken and P. L. Hammer. Linear separation of dominating sets in graphs. Ann.
Discrete Math., 3:1–10, 1978. Advances in graph theory (Cambridge Combinatorial
Conf., Trinity College, Cambridge, 1977).
[5] C. Berge and P. Duchet. Une généralisation du théorème de Gilmore. Cahiers Centre
Études Recherche Opér., 17(2-4):117–123, 1975. Colloque sur la Théorie des Graphes
(Paris, 1974).
[6] Claude Berge. Hypergraphs. North-Holland Publishing Co., Amsterdam, 1989.
[7] Louis J. Billera. Clutter decomposition and monotonic Boolean functions. Ann.
New York Acad. Sci., 175:41–48, 1970. International Conference on Combinatorial
Mathematics (1970).

the electronic journal of combinatorics 26(3) (2019), #P3.18 24

 [8] Louis J. Billera. On the composition and decomposition of clutters. J. Combinatorial
Theory Ser. B, 11:234–245, 1971.
[9] Endre Boros, Vladimir Gurvich, and Martin Milanič. On equistable, split, CIS,
and related classes of graphs. Discrete Appl. Math., 216(part 1):47–66, 2017.
doi:10.1016/j.dam.2015.07.023.
[10] Endre Boros, Vladimir Gurvich, and Martin Milanič. Characterizing and decom-
posing classes of threshold, split, and bipartite graphs via 1-Sperner hypergraphs.
CoRR, abs/1805.03405, 2018. arXiv:1805.03405.
[11] Gordon H. Bradley. Transformation of integer programs to knapsack problems. Dis-
crete Math., 1(1):29–45, 1971/72.
[12] Nina Chiarelli and Martin Milanič. Linear separation of connected dominating sets
in graphs (extended abstract). Proceedings of ISAIM 2014, International Symposium
on Artificial Intelligence and Mathematics, Fort Lauderdale, FL. January 6–8, 2014.
Available on the conference web page, http://www.cs.uic.edu/Isaim2014/.
[13] Nina Chiarelli and Martin Milanič. Total domishold graphs: a generalization of
threshold graphs, with connections to threshold hypergraphs. Discrete Appl. Math.,
179:1–12, 2014. doi:10.26493/1855-3974.1330.916.
[14] Nina Chiarelli and Martin Milanič. Linear separation of connected dominating sets
in graphs. Ars Math. Contemp., 16:487–525, 2019. doi:j.dam.2014.09.001.
[15] C.K. Chow. Boolean functions realizable with single threshold devices. In Proceedings
of the IRE, 39, pages 370–371, 1961.
[16] Václav Chvátal and Peter L. Hammer. Aggregation of inequalities in integer pro-
gramming. In Studies in integer programming (Proc. Workshop, Bonn, 1975), pages
145–162. Ann. of Discrete Math., Vol. 1. North-Holland, Amsterdam, 1977.
[17] Yves Crama and Peter L. Hammer, editors. Boolean Models and Methods in
Mmathematics, Computer Science, and Engineering, volume 134 of Encyclopedia of
Mathematics and its Applications. Cambridge University Press, Cambridge, 2010.
doi:10.1017/CBO9780511780448.
[18] Yves Crama and Peter L. Hammer. Boolean Functions: Theory, Algorithms, and
Applications, volume 142 of Encyclopedia of Mathematics and its Applications. Cam-
bridge University Press, Cambridge, 2011. doi:10.1017/CBO9780511852008.
[19] Jack Edmonds and D. R. Fulkerson. Bottleneck extrema. J. Combinatorial Theory,
8:299–306, 1970.
[20] Calvin C. Elgot. Truth functions realizable by single threshold organs. In Proceedings
of the Second Annual Symposium on Switching Circuit Theory and Logical Design,
1961 (SWCT 1961)., pages 225–245, Oct 1961. doi:10.1109/FOCS.1961.39.
[21] S. E. Elmaghraby and M. K. Wig. On the treatment of stock cutting problems as
diophantine programs. 1970. Operations Research Report No. 61, North Carolina
State University at Raleigh.

the electronic journal of combinatorics 26(3) (2019), #P3.18 25

[22] D. R. Fulkerson. Networks, frames, blocking systems. In Mathematics of the Decision
Sciences, Part 1 (Seminar, Stanford, Calif., 1967), pages 303–334. Amer. Math. Soc.,
Providence, R.I., 1968.
[23] F. Glover and R. E. D. Woolsey. Aggregating Diophantine equations. Z. Operations
Res. Ser. A-B, 16:A1–A10, 1972.
[24] Martin Charles Golumbic. Algorithmic Graph Theory and Perfect Graphs. Aca-
demic Press [Harcourt Brace Jovanovich, Publishers], New York-London-Toronto,
Ont., 1980.
[25] Ephraim Korach, Uri N. Peled, and Udi Rotics. Equistable distance-hereditary
graphs. Discrete Appl. Math., 156(4):462–477, 2008.
[26] S. L. Lauritzen, T. P. Speed, and K. Vijayan. Decomposable graphs and hypergraphs.
J. Austral. Math. Soc. Ser. A, 36(1):12–29, 1984.
[27] Vadim E. Levit and Martin Milanič. Equistable simplicial, very well-
covered, and line graphs. Discrete Appl. Math., 165:205–212, 2014.
doi:10.1016/j.dam.2013.01.022.
[28] N. V. R. Mahadev and U. N. Peled. Threshold Graphs and Related Topics, volume 56
of Annals of Discrete Mathematics. North-Holland Publishing Co., Amsterdam, 1995.
[29] N.V.R Mahadev, Uri N. Peled, and Feng Sun. Equistable graphs. J. Graph Theory,
18(3):281–299, 1994.
[30] G. B. Mathews. On the partition of numbers. Proc. Lond. Math. Soc., 28:486–490,
1896/97.
[31] Štefko Miklavič and Martin Milanič. Equistable graphs, general partition graphs,
triangle graphs, and graph products. Discrete Appl. Math., 159(11):1148–1159, 2011.
[32] Martin Milanič and Nicolas Trotignon. Equistarable graphs and counterexamples
to three conjectures on equistable graphs. J. Graph Theory, 84(4):536–551, 2017.
doi:10.1002/jgt.22040.
[33] Saburo Muroga. Threshold Logic and its Applications. Wiley-Interscience [John
Wiley & Sons], New York-London-Sydney, 1971.
[34] Manfred W. Padberg. Equivalent knapsack-type formulations of bounded integer
linear programs: An alternative approach. Naval Res. Logist. Quart., 19:699–708,
1972.
[35] Charles Payan. A class of threshold and domishold graphs: equistable and equidom-
inating graphs. Discrete Math., 29(1):47–52, 1980.
doi:10.1016/0012-365X(90)90286-Q.
[36] Uri N. Peled and Udi Rotics. Equistable chordal graphs. Discrete Appl. Math., 132:
203–210, 2003.
[37] Uri N. Peled and Bruno Simeone. Polynomial-time algorithms for regular set-covering
and threshold synthesis. Discrete Appl. Math., 12(1):57–69, 1985.
doi:10.1016/0166-218X(85)90040-X.

the electronic journal of combinatorics 26(3) (2019), #P3.18 26

[38] Jan Reiterman, Vojtěch Rödl, Edita Šiňajová, and Miroslav Tůma Threshold hyper-
graphs. Discrete Math., 54(2):193–200, 1985. doi:10.1016/0012-365X(85)90080-9.
[39] I. G. Rosenberg. Aggregation of equations in integer programming. Discrete Math.,
10:325–341, 1974.
[40] Harold N. Shapiro. On the counting problem for monotone boolean functions. Comm.
Pure Appl. Math., 23:299–312, 1970. doi:10.1002/cpa.3160230305.
[41] Jan-Georg Smaus. On Boolean functions encodable as a single linear pseudo-Boolean
constraint. In Pascal Van Hentenryck and Laurence Wolsey, editors, Integration of
AI and OR Techniques in Constraint Programming for Combinatorial Optimization
Problems, pages 288–302, Berlin, Heidelberg, 2007. Springer Berlin Heidelberg. ISBN
978-3-540-72397-4.
[42] Emanuel Sperner. Ein Satz über Untermengen einer endlichen Menge. Math. Z., 27
(1):544–548, 1928. doi:10.1007/BF01171114.
[43] Nikolaı̆ Nikolaevich Vorob’ev. Coalition games. Teor. Verojatnost. i Primenen., 12:
289–306, 1967.

Appendix: Proof of Theorem 27

Theorem 27 (restated). Every 1-Sperner hypergraph is equilizable.
Proof. We will show by induction on n = |V | that for every 1-Sperner hypergraph H =
(V, E) there exists a positive integer weight function w : V → Z>0 and a non-negative
integer threshold t ∈ Z>0 such that for every subset X ⊆ V , we have w(X) = t if and
only if X ∈ E. This will establish the equilizability of H.
For n = 0, we can take the (empty) mapping given by w(x) = 1 for all x ∈ V and the
threshold 
1, if E = ∅;
t=
0, if E = {∅}.

Now, let H = (V, E) be a 1-Sperner hypergraph with n > 1. By Theorem 24, H is

a safe gluing of two 1-Sperner hypergraphs, say H = H1 H2 with H1 = (V1 , E1 ) and
H2 = (V2 , E2 ), where V = V1 ∪ V2 ∪ {z}, V1 ∩ V2 = ∅, and z 6∈ V1 ∪ V2 . By the inductive
hypothesis, H1 and H2 are equilizable, that is, there exist positive integer weight functions
wi : Vi → Z>0 and non-negative integer thresholds ti ∈ Z>0 for i = 1, 2 such that for every
subset X ⊆ Vi , we have wi (X) = ti if and only if X ∈ Ei .
Let us define the threshold t = M w1 (V1 ) + t2 , where M = w2 (V2 ) + 1, and the weight
function w : V → Z>0 by the rule

 M w1 (x), if x ∈ V1 ;
w(x) = w2 (x), if x ∈ V2 ;

M (w1 (V1 ) − t1 ) + t2 , if x = z.
Since the weight function defined above coincides with the one in the proof of Theorem 26,
this function is indeed strictly positive.

the electronic journal of combinatorics 26(3) (2019), #P3.18 27

 We claim that for every subset X ⊆ V , we have w(X) = t if and only if X ∈ E. This
will establish the equilizability of H.
Suppose first that w(X) = t for some X ⊆ V . Let Xi = X ∩ Vi for i = 1, 2. For later
use, we note that
w2 (X2 ) 6 w2 (V2 ) < M . (4)
We consider two cases depending on whether z ∈ X or not. Suppose first that z ∈ X.
Then
M w1 (V1 ) + t2 = t = w(X)
= w(z) + w(X1 ) + w(X2 )
= M (w1 (V1 ) − t1 ) + t2 + M w1 (X1 ) + w2 (X2 ) ,
which implies
M w1 (X1 ) + w2 (X2 ) = M t1 . (5)
Using (4), we obtain M w1 (X1 )+w2 (X2 ) = M t1 > M (t1 −1)+w2 (X2 ) and hence w1 (X1 ) >
t1 . Moreover, if w1 (X1 ) > t1 + 1, then M w1 (X1 ) + w2 (X2 ) > M t1 + M + w2 (X2 ) > M t1 ,
contradicting (5). We infer that w1 (X1 ) = t1 and consequently X1 ∈ E1 . Equation (5)
together with w1 (X1 ) = t1 implies that w2 (X2 ) = 0. Since w2 is positive on all V2 , it
follows that X2 = ∅. Therefore, we have X = {z} ∪ X1 ∈ E.
Now, suppose that z 6∈ X. In this case, M w1 (V1 ) + t2 = t = w(X) = w(X1 ) + w(X2 ) =
M w1 (X1 ) + w2 (X2 ), which implies
M w1 (X1 ) + w2 (X2 ) = M w1 (V1 ) + t2 . (6)
We must have X1 = V1 since if there exists a vertex v ∈ V1 \ X1 , then we would have
M w1 (X1 ) + w2 (X2 ) 6 M w1 (V1 ) − M w1 (v) + w2 (X2 )
6 M w1 (V1 ) − M + w2 (X2 )
< M w1 (V1 ) + t2 ,
where the last inequality follows from (4) and t2 > 0. Therefore, equality (6) simplifies
to w2 (X2 ) = t2 , and consequently there exists a hyperedge X2 ∈ E2 . This implies that
X = V1 ∪ X2 ∈ E.
For the converse direction, suppose that X is a subset of V such that X ∈ E. We
need to show that w(X) = t. We again consider two cases depending on whether z ∈ X
or not. Suppose first that z ∈ X. Then X = {z} ∪ X1 for some X1 ∈ E1 . Due to the
property of w1 , we have w1 (X1 ) = t1 . Consequently,
w(X) = w(z) + w(X1 )
= M (w1 (V1 ) − t1 ) + t2 + M w1 (X1 )
= M w1 (V1 ) − M t1 + t2 + M t1
= M w1 (V1 ) + t2 = t .
Suppose now that z 6∈ X. Then X = V1 ∪ X2 for some X2 ∈ E2 . Due to the property of
w2 , we have w2 (X2 ) = t2 . It follows that w(X) = w(V1 ) + w(X2 ) = M w1 (V1 ) + w2 (X2 ) =
M w1 (V1 ) + t2 = t. This completes the proof.

the electronic journal of combinatorics 26(3) (2019), #P3.18 28