def test_stacked_feature_importance(mock_fit, X_y_regression,
stackable_regressors):
X, y = X_y_regression
input_pipelines = [
RegressionPipeline([regressor]) for regressor in stackable_regressors
]
clf = StackedEnsembleRegressor(input_pipelines=input_pipelines, n_jobs=1)
clf.fit(X, y)
mock_fit.assert_called()
clf._is_fitted = True
with pytest.raises(NotImplementedError,
match="feature_importance is not implemented"):
clf.feature_importance
def test_stacked_ensemble_multilevel(linear_regression_pipeline_class):
# checks passing a stacked ensemble classifier as a final estimator
X = pd.DataFrame(np.random.rand(50, 5))
y = pd.Series(np.random.rand(50, ))
base = StackedEnsembleRegressor(
input_pipelines=[linear_regression_pipeline_class(parameters={})],
n_jobs=1)
clf = StackedEnsembleRegressor(
input_pipelines=[linear_regression_pipeline_class(parameters={})],
final_estimator=base,
n_jobs=1)
clf.fit(X, y)
y_pred = clf.predict(X)
assert len(y_pred) == len(y)
assert not np.isnan(y_pred.to_series()).all()
def test_stacked_ensemble_n_jobs_negative_one(
X_y_regression, linear_regression_pipeline_class):
X, y = X_y_regression
input_pipelines = [linear_regression_pipeline_class(parameters={})]
clf = StackedEnsembleRegressor(input_pipelines=input_pipelines)
expected_parameters = {
"input_pipelines": input_pipelines,
"final_estimator": None,
'cv': None,
'n_jobs': -1
}
assert clf.parameters == expected_parameters
clf.fit(X, y)
y_pred = clf.predict(X)
assert len(y_pred) == len(y)
assert not np.isnan(y_pred.to_series()).all()
def test_stacked_fit_predict_regression(X_y_regression, stackable_regressors):
X, y = X_y_regression
input_pipelines = [
RegressionPipeline([regressor]) for regressor in stackable_regressors
]
clf = StackedEnsembleRegressor(input_pipelines=input_pipelines, n_jobs=1)
clf.fit(X, y)
y_pred = clf.predict(X)
assert len(y_pred) == len(y)
assert isinstance(y_pred, ww.DataColumn)
assert not np.isnan(y_pred.to_series()).all()
clf = StackedEnsembleRegressor(input_pipelines=input_pipelines,
final_estimator=RandomForestRegressor(),
n_jobs=1)
clf.fit(X, y)
y_pred = clf.predict(X)
assert len(y_pred) == len(y)
assert isinstance(y_pred, ww.DataColumn)
assert not np.isnan(y_pred.to_series()).all()
def test_stacked_ensemble_init_with_multiple_same_estimators(
X_y_regression, linear_regression_pipeline_class):
# Checks that it is okay to pass multiple of the same type of estimator
X, y = X_y_regression
input_pipelines = [
linear_regression_pipeline_class(parameters={}),
linear_regression_pipeline_class(parameters={})
]
clf = StackedEnsembleRegressor(input_pipelines=input_pipelines, n_jobs=1)
expected_parameters = {
"input_pipelines": input_pipelines,
"final_estimator": None,
'cv': None,
'n_jobs': 1
}
assert clf.parameters == expected_parameters
fitted = clf.fit(X, y)
assert isinstance(fitted, StackedEnsembleRegressor)
y_pred = clf.predict(X)
assert len(y_pred) == len(y)
assert not np.isnan(y_pred.to_series()).all()