Giters

ragibson / ModularityPruning

Pruning tool to identify small subsets of network partitions that are significant from the perspective of stochastic block model inference. This method works for single-layer and multi-layer networks, as well as for restricting focus to a fixed number of communities when desired.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Reinvestigate recent Python parallelism / multiprocessing improvements

ragibson opened this issue · comments

commented

I recall our original implementation of the parallel optimization functions having to work around memory leaks in Python's multiprocessing package, so we should check if those are still needed.

More importantly, we should check that the parallelism is working well for small cases since I've learned some users are running this on very small networks. This may also warrant basic performance regression checks in our test suites.

commented

This is primarily addressed in 41e1cc2 and 81b287d.

commented

Closing now that the automated performance scaling test has passed on two of my machines and the GitHub action.