Categorical Pipeline

Wyx Smrf

Pipeline

The code creates a pipeline using the SimpleImputer() algorithm and the OneHotEncoder() transformer. The pipeline has a preprocessing column that will include any values that are not encoder handles that are unknown.

Library: Scikit-Learn

Filename pattern: .ipynb, .py

from sklearn.impute import SimpleImputer
from sklearn.preprocessing import OneHotEncoder
from sklearn.compose import make_column_transformer, make_column_selector

cat_pipeline = make_pipeline(
  SimpleImputer(strategy="most_frequent"),
  OneHotEncoder(handle_unknown="ignore"))

preprocessing = make_column_transformer((cat_pipeline, make_column_selector(dtype_include=object)))