AttributeError: 'Ridge' object has no attribute 'feature_names_in_'

[21564669]

Describe the bug

the issue appears in the example of https://scikit-learn.org/stable/auto_examples/linear_model/plot_ridge_coeffs.html#sphx-glr-auto-examples-linear-model-plot-ridge-coeffs-py

in the following piece of code, if we add 'print(f"clf.feature_names_in:{clf.feature_names_in_}")' after the fit() function is called,
it will pop up an issue that 'AttributeError: 'Ridge' object has no attribute 'feature_names_in_''

# Train the model with different regularisation strengths
for a in alphas:
    clf.set_params(alpha=a)
    clf.fit(X, y)
    print(f"clf.feature_names_in:{clf.feature_names_in_}")
    coefs.append(clf.coef_)
    errors.append(mean_squared_error(clf.coef_, w))

sklearn version is 1.01

Steps/Code to Reproduce

import matplotlib.pyplot as plt
import numpy as np

from sklearn.datasets import make_regression
from sklearn.linear_model import Ridge
from sklearn.metrics import mean_squared_error

clf = Ridge()

X, y, w = make_regression(
    n_samples=10, n_features=10, coef=True, random_state=2, bias=0.0
)
coefs = []
errors = []

alphas = np.logspace(-6, 6, 200)
for a in alphas:
    clf.set_params(alpha=a)
    clf.fit(X, y)
    print(f"clf.feature_names_in:{clf.feature_names_in_}")
    coefs.append(clf.coef_)
    errors.append(mean_squared_error(clf.coef_, w))

Expected Results

it is expected to print the attribute of feature_names_in_

Actual Results

but it raised an error.
AttributeError: 'Ridge' object has no attribute 'feature_names_in_'

Versions

System:
python: 3.8.0 (tags/v3.8.0:fa919fd, Oct 14 2019, 19:37:50) [MSC v.1916 64 bit (AMD64)]
executable: C:\Users\ezqiuya\AppData\Local\Programs\Python\Python38\python.exe
machine: Windows-10-10.0.18362-SP0

Python dependencies:
pip: 21.3.1
setuptools: 59.1.1
sklearn: 1.0.1
numpy: 1.21.4
scipy: 1.7.2
Cython: 0.29.24
pandas: 1.3.4
matplotlib: 3.5.0
joblib: 1.1.0
threadpoolctl: 3.0.0

Built with OpenMP: True

[glemaitre]

As documented in the attributes section of the Ridge documentation (and this rule apply to all estimator), feature_names_in_ is only available if the X as all string columns:

In [1]: import matplotlib.pyplot as plt
   ...: import numpy as np
   ...: import pandas as pd
   ...: from sklearn.datasets import make_regression
   ...: from sklearn.linear_model import Ridge
   ...: from sklearn.metrics import mean_squared_error
   ...: 
   ...: clf = Ridge()
   ...: 
   ...: X, y, w = make_regression(
   ...:     n_samples=10, n_features=10, coef=True, random_state=2, bias=0.0
   ...: )
   ...: X = pd.DataFrame(X, columns=[f"Features {i}" for i in range(X.shape[1])])
   ...: coefs = []
   ...: errors = []
   ...: 
   ...: alphas = np.logspace(-6, 6, 200)
   ...: for a in alphas:
   ...:     clf.set_params(alpha=a)
   ...:     clf.fit(X, y)
   ...:     print(f"clf.feature_names_in:{clf.feature_names_in_}")
   ...:     coefs.append(clf.coef_)
   ...:     errors.append(mean_squared_error(clf.coef_, w))
   ...: 

In your case, a NumPy array has no column names so you could generate the column name with range(X.shape[1]).