Three Columns That Count

A scaled logistic regression classifies the three wine growers. Which columns does it actually lean on, measured on wines it has not seen?

Task: write top_columns(n).

• Load load_wine(return_X_y=True, as_frame=True).
• Split with train_test_split(X, y, test_size=0.3, random_state=1, stratify=y).
• Fit make_pipeline(StandardScaler(), LogisticRegression(max_iter=1000)) on the training rows.
• Run permutation_importance(model, X_test, y_test, n_repeats=10, random_state=0).
• Return the names of the n columns with the highest mean importance, highest first.