Tools for experiments on metrics used for matching.
apxgi.py runs MCMC to estimate the distribution of node correctness for a given edge correctness.
To generate results for synthetic graphs:
python run.py 500 0.03 ER noperturb
python run.py 500 0.06 ER noperturb
python run.py 500 0.09 ER noperturb
python run.py 500 0.004 BA noperturb
python run.py 500 0.01 BA noperturb
python run.py 500 0.03 BA noperturb
python run.py 500 0.06 BA noperturb
python run.py 500 0.09 BA noperturb
python run.py 500 0.0001 WS noperturb
python run.py 500 0.001 WS noperturb
python run.py 500 0.01 WS noperturb
python run.py 500 0.05 WS noperturb
python run.py 500 0.02 GEO noperturb
python run.py 500 0.03 GEO noperturb
python run.py 500 0.06 GEO noperturb
python run.py 500 0.09 GEO noperturb
python run.py 500 0.1 EV noperturb
python run.py 500 0.24 EV noperturb
python run.py 500 0.0 SL noperturb
To generate perturbed results:
python run.py 500 0.03 ER thin 0.25
(etc for scramble, rewire, randomize)
To generate results comparing different graph sampling methods:
python sampRun.py 500 0.02 GEO
python sampRun.py 500 0.03 GEO
python sampRun.py 500 0.06 GEO
python sampRun.py 500 0.09 GEO
python sampRun.py 500 0.0 SL
cd samprun/noperturb
python ../../sampPlots.py
To generate results using XS sampling on PPI graphs:
python run.py 500 0 PPI noperturb 0 human
python run.py 500 0 PPI noperturb 0 fly
python run.py 500 0 PPI noperturb 0 mouse
python run.py 500 0 PPI noperturb 0 worm
python run.py 500 0 PPI noperturb 0 yeast
python run.py 500 0 PPI scramble 0.25 human
python run.py 500 0 PPI scramble 0.25 fly
python run.py 500 0 PPI scramble 0.25 mouse
python run.py 500 0 PPI scramble 0.25 worm
python run.py 500 0 PPI scramble 0.25 yeast