Generate random data for scalability analysis

[koksal]

Purpose
Generate random graphs with maximum given degree, and associated time series node, to assess solver performance on problems of different size.

Technical plan

• Generate random graphs.
  • Generate fully random graphs.
  • Generate graphs by starting from the source nodes of an existing graph and randomly adding edges without exceeding a node degree limit.
  • Efficient random graph generation from existing source graph.
• Generate time series (and significance) data on a given graph.
• Measure solver performance on networks of different size.
• Adjust data characteristics in random graph generation to real data.
  • Add graph stats.
  • Add time series data coverage stats.
  • Performance measurement replicates.

[koksal]

Simple statistics from actual data:

Expanded graph stats:
# vertices: 376
# edges   : 724
Average vertex degree: 14.55
Median vertex degree : 10.5
Ratio of vertices with data: 0.6196808510638298

[solver] Elapsed time: 0.805911922s