Runtime warning from scipy

[marcenko]

Running this JuGEx analysis:

from os import environ
environ['HBP_AUTH_TOKEN'] = "<<<Paste your HBP auth token here>>>"

import brainscapes as bs
bs.logger.setLevel("INFO")

atlas = bs.atlases.MULTILEVEL_HUMAN_ATLAS
atlas.select_parcellation(bs.parcellations.JULICH_BRAIN_PROBABILISTIC_CYTOARCHITECTONIC_MAPS_V2_5_)
atlas.enable_continuous_map_thresholding(0.2)

candidate_regions = ["v1 right", "v2 right"]
candidate_genes = ["MAOA","TAC1"]

jugex = bs.analysis.DifferentialGeneExpression(atlas)
jugex.add_candidate_genes(candidate_genes)
jugex.define_roi1(candidate_regions[0])
jugex.define_roi2(candidate_regions[1])

result = jugex.run(permutations=1000)
print(result['p-values'])

Causes this runtime warning from scipy:

/Users/xxxx/opt/anaconda3/lib/python3.8/site-packages/scipy/stats/_distn_infrastructure.py:1932: RuntimeWarning: invalid value encountered in less_equal
  cond2 = cond0 & (x <= _a)

Versions:

• OS: macOS 10.15.5
• Python 3.8.3
• scipy: 1.5.0

Is anybody else experiencing similar warnings?

---------- old comments ----------
Dickscheid, Timo
I cannot reproduce (Python 3.9.1, scipy 1.6.0). But I assume it must be in the winsorize() call of DifferentialGeneAnalysis._retrieve_samples(). Could you dump the inputs to the method call, and verify?

Well, actually you just need to dump the GeneExpression features, winsorize averages the z-scores of the extracted GeneExpression features.