Error create_shap_features (task: binary)

[whatsdog0001]

task is binary but error on multiclass

---

TypeError Traceback (most recent call last)
in <cell line: 8>()
6 X_valid = X_valid.drop(columns=['race_id', 'rank'])
7
----> 8 selected_features_df, shap_features = shap_select(
9 model, X_valid, y_valid, task='binary', threshold=0.05, return_extended_data=True
10 )

in shap_select(tree_model, validation_df, target, feature_names, task, threshold, return_extended_data, alpha)
316 )
317 else:
--> 318 shap_features = create_shap_features(tree_model, validation_df[feature_names])
319
320 # Compute statistical significance of each feature, recursively ablating

in create_shap_features(tree_model, validation_df, classes)
37 elif len(shap_values.shape) == 3: # multiclass classification
38 out = {}
---> 39 for i, c in enumerate(classes):
40 out[i] = pd.DataFrame(
41 shap_values[:, :, i],

TypeError: 'NoneType' object is not iterable

[whatsdog0001]

Fixed

pip install -U lightgbm
pip install -U shap