Visualizing Feature Importances with PermutationImportance in Eli5

This function uses Eli5's PermutationImportance to evaluate feature importance and visualizes the results through eli5.show_weights.

Technology Stack : Eli5, PermutationImportance

Code Type : Python Function

Code Difficulty : Intermediate

                
                    
def visualize_feature_importances(model, X, feature_names):
    """
    Visualize feature importances from a fitted model using Eli5.

    Args:
        model: Fitted machine learning model.
        X: Input features.
        feature_names: Names of the input features.
    """
    import eli5
    from eli5.sklearn import PermutationImportance

    # Fit the PermutationImportance model
    perm = PermutationImportance(model, random_state=42).fit(X, y)

    # Create a permutation feature importance visualization
   eli5.show_weights(perm, feature_names=feature_names, top=10)