Closed miguelgondu closed 1 month ago
Hey @RMichae1,
This PR addresses the bug you noticed on foldx
parallelization. It ran on the poli__protein
env, but not in isolation. The problem was trying to pickle the connection to the isolated function, and that's fixed by having every children in the parallel call create its own connection.
It would be ideal to get your eyes on this (even if it's a brief pass), but feel free to merge without a full review.
This PR fixes a bug we had on parallel running of
foldx
(see the discussion in #184).In the process, I've also implemented a
get_inner_function
utility that abstracts away the ugly code repetition we had on every isolate-able black box.(Closes #184 and closes #172)