Closed EzryStIago closed 1 year ago
Tracked down to this line in processing.py:
grp_series = group.dropna(axis=1).iloc[:, use_col]
Resolved by:
grp_series = group.dropna(axis=1).sort_index(axis=1).iloc[:, use_col]
in commit 14e247ea3ff0fc9ddaf99bdf09cd61ddaa1d5c20
This is especially common in embarrassingly parallel runs. E.g. FEP_number_of_samples.pdf