def test_compare_model_classification():
x_train, y_train = make_classification(n_samples=50,
n_features=4,
n_informative=2,
n_redundant=0,
random_state=0,
shuffle=False)
model_list = [
RandomForestClassifier(n_estimators=5, max_depth=2, random_state=0),
GradientBoostingClassifier(n_estimators=5, max_depth=2, random_state=0)
]
fitted_model, model_scores = model.compare_model(model_list, x_train,
y_train, 'accuracy')
assert type(fitted_model) is list
assert type(model_scores) is list
assert hasattr(fitted_model[0], "predict")
def test_compare_model():
x_train, y_train = make_classification(
n_samples=50,
n_features=4,
n_informative=2,
n_redundant=0,
random_state=0,
shuffle=False
)
model_list = [
RandomForestClassifier(n_estimators=100, max_depth=2, random_state=0),
BaggingClassifier(),
GradientBoostingClassifier()
]
scoring_metrics = ['accuracy']
fitted_model, model_scores = model.compare_model(model_list, x_train, y_train, True, scoring_metrics)
assert type(fitted_model) is list
assert type(model_scores) is list
def test_compare_model_regression():
x_train, y_train = make_classification(n_samples=50,
n_features=4,
n_informative=2,
n_redundant=0,
random_state=0,
shuffle=False)
model_list = [
RandomForestRegressor(n_estimators=5, max_depth=2, random_state=0),
GradientBoostingRegressor(n_estimators=5, max_depth=2, random_state=0)
]
fitted_model, model_scores = compare_model(model_list,
x_train,
y_train,
'neg_mean_absolute_error',
plot=False)
assert type(fitted_model) is list
assert type(model_scores) is list
assert hasattr(fitted_model[0], "predict")