• cycle_detect.py:
  • cycle_detection(edges): yields networkx.DiGraph each time when encounter cycle
    • edges: [(source, target),]
    • time complexity: Õ(m√n)
• test.py:
  • command arg: #nodes #edges

This is an implementation based on the following published paper：

@inproceedings{Bernstein:2018:ITS:3174304.3175268,

author = {Bernstein, Aaron and Chechik, Shiri},

title = {Incremental Topological Sort and Cycle Detection in [Equation] Expected Total Time},

booktitle = {Proceedings of the Twenty-Ninth Annual ACM-SIAM Symposium on Discrete Algorithms},

series = {SODA '18},

year = {2018},

isbn = {978-1-6119-7503-1},

location = {New Orleans, Louisiana},

pages = {21--34},

numpages = {14},

url = {http://dl.acm.org/citation.cfm?id=3174304.3175268},

acmid = {3175268},

publisher = {Society for Industrial and Applied Mathematics},

address = {Philadelphia, PA, USA}, }