I was reading the scikitlearn tutorial about column transformer. The given example (https://scikit-learn.org/stable/modules/generated/sklearn.compose.make_column_selector.html#sklearn.compose.make_column_selector) works, but when I tried to select only few columns, It gives me error.

MWE

import numpy as np
import pandas as pd
import seaborn as sns

from sklearn.compose import make_column_transformer
from sklearn.compose import make_column_selector

df = sns.load_dataset('tips')
mycols = ['tip','sex']


ct = make_column_transformer(make_column_selector(pattern=mycols)
ct.fit_transform(df)

Required

I want only the select columns in the output.

NOTE
Of course, I know I can do df[mycols], I am looking for scikit learn pipeline example.

If you don't mind mlxtend, it has built-in transformer for that.

Using mlxtend

from mlxtend.feature_selection import ColumnSelector

pipe = ColumnSelector(mycols)
pipe.fit_transform(df)

For sklearn >= 0.20

• Reference: https://scikit-learn.org/stable/modules/generated/sklearn.compose.ColumnTransformer.html

from sklearn.compose import ColumnTransformer
from sklearn.pipeline import Pipeline
import seaborn as sns

df = sns.load_dataset('tips')
mycols = ['tip','sex']

pipeline = Pipeline([
    ("selector", ColumnTransformer([
        ("selector", "passthrough", mycols)
    ], remainder="drop"))
])

pipeline.fit_transform(df)

For sklearn < 0.20

from sklearn.base import BaseEstimator, TransformerMixin
from sklearn.pipeline import Pipeline

class FeatureSelector(BaseEstimator, TransformerMixin):
    def __init__(self, columns):
        self.columns = columns

    def fit(self, X, y=None):
        return self

    def transform(self, X, y=None):
        return X[self.columns]


pipeline = Pipeline([('selector', FeatureSelector(columns=mycols))
                     ])

pipeline.fit_transform(df)[:5]