Pipeline and GridSearchCV

Introduction

Create reproducible machine learning workflows with pipelines.

Pipeline Basics

from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.decomposition import PCA
from sklearn.linear_model import LogisticRegression

pipeline = Pipeline([
    ("scaler", StandardScaler()),
    ("pca", PCA(n_components=2)),
    ("classifier", LogisticRegression())
])

pipeline.fit(X_train, y_train)
predictions = pipeline.predict(X_test)

Pipeline with GridSearch

from sklearn.model_selection import GridSearchCV

pipeline = Pipeline([
    ("scaler", StandardScaler()),
    ("classifier", LogisticRegression())
])

param_grid = {
    "classifier__C": [0.1, 1, 10],
    "scaler": [StandardScaler(), None]
}

grid = GridSearchCV(pipeline, param_grid, cv=5)
grid.fit(X, y)

Column Transformer

from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import StandardScaler, OneHotEncoder

preprocessor = ColumnTransformer([
    ("num", StandardScaler(), numerical_features),
    ("cat", OneHotEncoder(), categorical_features)
])

Practice Problems

1. Build preprocessing pipeline
2. Combine with GridSearch
3. Use ColumnTransformer
4. Access pipeline steps
5. Create custom transformers