[sklearn] Sklearn #python #ml #sklearn

column_selector_in_pipeline.py

from sklearn.compose import ColumnTransformer, make_column_selector
from sklearn.pipeline import Pipeline

col_name_pattern = "feature_"
pl = Pipeline([
    (
        "col_selector",
        ColumnTransformer([
            ("selector", "passthrough", make_column_selector(pattern=col_name_pattern))
        ], remainder="drop")
    ),
])