[BUG]

[narminjabbarli]

Describe the bug

when I run feature_selector fit, I get index error, I'll put the screenshot below. Firstly, I supposed there was something wrong with my dataset, but later I tried with load data (classification)

Screenshots

[kaustv-datta]

I am also getting this error after updating my env to python 3.10. Things were fine in 3.9.

[104410]

I still got this error in py3.9，but it work in py3.8

[SUNgardener]

Issue Confirmation & Solution Proposal

Issue Reproduction:
I've successfully reproduced the error described in this issue. The root cause stems from incorrect axis selection when calculating mean SHAP values, leading to dimension mismatch in subsequent operations.

Proposed Fix:
In src/BorutaShap.py:

# Original (Lines 843 & 867)
self.shap_values = self.shap_values.mean(0)

# Modified
self.shap_values = self.shap_values.mean(1)  # Axis correction