def test_data(df,model, fit=False):
target = df.target
features = df.drop("target", axis=1)
X_train, X_test, y_train, y_test = train_test_split(features,target)
if fit:
preds =model.predict(X_test)
else:
preds,_=create_model(model,X_train, X_test, y_train)
evaluate_model(preds,y_test)
def test_model(df):
gbc= GradientBoostingClassifier()
test_data(df, gbc)
gbc_params = {
'min_samples_leaf': [1, 5, 10, 20],
'n_estimators': [100,200,300,400],
'learning_rate': [0.01, 0.025, 0.05, 0.1],
}
search = GridSearchCV(gbc, gbc_params, n_jobs=-1, scoring = "f1")
search.fit(X_train, y_train)
preds_b = search.best_estimator_.predict(X_test)
evaluate_model(preds_b, y_test)
return search.best_estimator_
