aggregate_importances

aggregate_importances#

mlproject.postprocess.feature_importances.aggregate_importances(dfs, value_col, all_features)[source]#

Aggregate feature importance DataFrames across folds.

Parameters:

  • dfs (list[pd.DataFrame]) – List of DataFrames containing feature importances.

  • value_col (str) – Column name for importance values to aggregate.

  • all_features (list[str]) – List of all features to ensure consistent indexing.

Returns:

df_agg – Aggregated DataFrame with mean and std of importances.

Return type:

pd.DataFrame

Contents