Describe the workflow you want to enable

I would like to get the feature names input to a classifier after imputation.

column_trans = ColumnTransformer([('imputed', SimpleImputer(add_indicator=True), x_con)], remainder='drop')
# column_trans = ColumnTransformer([('scaler', preprocessing.StandardScaler(), x_con)], remainder='drop')
pipe = make_pipeline(column_trans, LogisticRegressionCV(cv=2, random_state=0))
pipe.fit(X, y)
pipe[:-1].get_feature_names_out()

Describe your proposed solution

Implement get_feature_names_out for SimpleImputer as it is implemented for other transformers e.g. it works for StandardScaler.

Describe alternatives you've considered, if relevant

No response

Additional context

No response