scikit-learn 'ValueError: Expected 2D array, got 1D array instead: array=[...]. Reshape your data either using array.reshape(-1, 1) if your data has a single feature or array.reshape(1, -1) if it contains a single sample.' How do I fix it?

Cause: scikit-learn estimators expect X (the features) as a 2-D array shaped (n_samples, n_features): one row per sample, one column per feature. Passing a 1-D array or a pandas Series (shape (n,)) is ambiguous: it could be n samples with one feature or one sample with n features, so fit, predict and transform refuse it. It commonly happens with a single feature column (df['x'] instead of df[['x']]) or when predicting a single sample. Fix: 1. One feature, many samples: make it a column: X = x.reshape(-1, 1) for NumPy (the -1 lets NumPy compute the number of rows), or select a DataFrame with double brackets: X = df[['feature']]. 2. One sample, many features (e.g. predicting one row): make it a row: model.predict(x.reshape(1, -1)), or pass a one-row DataFrame with the same column names used for training. 3. Keep y (the target) 1-D: y = df['target'] is correct; a column-shaped y can trigger a DataConversionWarning instead. 4. Check shapes before fitting: print(X.shape, y.shape) should show (n_samples, n_features) and (n_samples,). 5. When you trained on a DataFrame, predict with a DataFrame that has the same columns in the same order, to avoid feature-name warnings and silent column mix-ups. Verify: X.ndim == 2 and model.fit(X, y) and model.predict(X_new) run without the error. Sources: https://scikit-learn.org/stable/getting_started.html https://numpy.org/doc/stable/reference/generated/numpy.reshape.html