Example of a gridsearch in scikit-learn

gridsearchcv.py

from sklearn.model_selection import GridSearchCV

# params is a dictionary, the keys are the hyperparameter and the vaules are a list of values
# to search over.
params = [
    {
        "transform__txt__max_features": [None, 100, 10],
        "transform__num__selector__attribute_names": [
            ["n_words"],
            ["mean_word_length"],
            ["n_words", "mean_word_length"],
        ],
    }
]

# GridSearchCV by default stratifies our cross-validation
# and retrains model on the best set of hyperparameters
model = GridSearchCV(pipeline, params, scoring="balanced_accuracy", cv=5)
model.fit(X_train, y_train)

# Display all of the results of the grid search
print(model.cv_results_)
# Display the mean scores for each combination of hyperparameters
print(model.cv_results_[model.cv_results_["mean_test_score"]])