def create_iris_pandas():
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
X_train = pd.DataFrame(X_train, columns=feature_names)
X_test = pd.DataFrame(X_test, columns=feature_names)
return X_train, X_test, y_train, y_test, feature_names
def test_surrogate_error_tree_iris(self, analyzer_type):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
models = create_models_classification(X_train, y_train)
for model in models:
run_error_analyzer(model, X_test, y_test, feature_names,
analyzer_type)
def test_importances_iris(self):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
models = create_models_classification(X_train, y_train)
for model in models:
categorical_features = []
run_error_analyzer(model, X_test, y_test, feature_names,
categorical_features)
def test_matrix_filter_iris(self):
x_train, x_test, y_train, y_test, feature_names, _ = create_iris_data()
models = create_models(x_train, y_train)
for model in models:
categorical_features = []
run_error_analyzer(model, x_test, y_test, feature_names,
categorical_features)
def iris():
x_train, x_test, y_train, y_test, features, classes = create_iris_data()
yield {
DatasetConstants.X_TRAIN: x_train,
DatasetConstants.X_TEST: x_test,
DatasetConstants.Y_TRAIN: y_train,
DatasetConstants.Y_TEST: y_test,
DatasetConstants.FEATURES: features,
DatasetConstants.CLASSES: classes
}
def test_error_report_iris(self, alter_feature_names):
X_train, X_test, y_train, y_test, feature_names, _ = \
create_iris_data(append_special_characters=alter_feature_names)
models = create_models_classification(X_train, y_train)
for model in models:
categorical_features = []
run_error_analyzer(model, X_test, y_test, feature_names,
categorical_features,
expect_user_warnings=alter_feature_names)
def test_matrix_filter_iris_quantile_binning(self):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
model_task = ModelTask.CLASSIFICATION
run_error_analyzer_on_models(X_train,
y_train,
X_test,
y_test,
feature_names,
model_task,
quantile_binning=True)
def test_error_report_iris_numpy_int64_features(self):
X_train, X_test, y_train, y_test, _, _ = create_iris_data()
# Test with numpy feature indexes instead of string feature names
feature_names = range(0, X_train.shape[1])
feature_names = [np.int64(i) for i in feature_names]
models = create_models_classification(X_train, y_train)
for model in models:
categorical_features = []
run_error_analyzer(model, X_test, y_test, feature_names,
categorical_features)
def test_raianalyzer_iris(self):
x_train, x_test, y_train, y_test, feature_names, classes = \
create_iris_data()
x_train = pd.DataFrame(x_train, columns=feature_names)
x_test = pd.DataFrame(x_test, columns=feature_names)
models = create_models(x_train, y_train)
x_train[LABELS] = y_train
x_test[LABELS] = y_test
for model in models:
run_raianalyzer(model, x_train, x_test, LABELS, classes)
def test_matrix_filter_iris_num_bins(self):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
model_task = ModelTask.CLASSIFICATION
num_bins_list = [2, 4, 10, 12]
for num_bins in num_bins_list:
run_error_analyzer_on_models(X_train,
y_train,
X_test,
y_test,
feature_names,
model_task,
num_bins=num_bins)
def test_explain_model_multiclass_classification_with_different_format_predictions(
self, mimic_explainer, if_predictions_as_dataframe, explainable_model):
x_train, x_test, y_train, y_test, _, classes = create_iris_data()
model = LogisticRegression(random_state=42).fit(x_train, y_train)
model.fit(x_train, y_train)
model = PredictAsDataFrameClassificationTestModel(
model, return_predictions_as_dataframe=if_predictions_as_dataframe)
kwargs = {}
explainer = mimic_explainer(model, x_train, explainable_model, **kwargs)
global_explanation = explainer.explain_global(evaluation_examples=x_test)
assert global_explanation is not None
def test_model_analysis_iris(self, manager_type):
x_train, x_test, y_train, y_test, feature_names, classes = \
create_iris_data()
x_train = pd.DataFrame(x_train, columns=feature_names)
x_test = pd.DataFrame(x_test, columns=feature_names)
models = create_models_classification(x_train, y_train)
x_train[LABELS] = y_train
x_test[LABELS] = y_test
manager_args = {DESIRED_CLASS: 0}
for model in models:
run_model_analysis(model, x_train, x_test, LABELS, [],
manager_type, manager_args, classes)
def test_parameters(self, metric, min_child_samples, max_depth, num_leaves,
analyzer_type):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
model = create_kneighbors_classifier(X_train, y_train)
run_error_analyzer(model,
X_test,
y_test,
feature_names,
analyzer_type,
max_depth=max_depth,
num_leaves=num_leaves,
min_child_samples=min_child_samples,
metric=metric)
def test_matrix_filter_iris_invalid_num_bins(self):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
model_task = ModelTask.CLASSIFICATION
invalid_num_bins_list = [-10, -1, 0]
err = 'Number of bins parameter must be greater than 0 for the heatmap'
for num_bins in invalid_num_bins_list:
with pytest.raises(ValueError, match=err):
run_error_analyzer_on_models(X_train,
y_train,
X_test,
y_test,
feature_names,
model_task,
num_bins=num_bins)
def test_matrix_filter_iris_int64(self):
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
X_train = pd.DataFrame(X_train, columns=feature_names)
X_test = pd.DataFrame(X_test, columns=feature_names)
X_train[feature_names[0]] = X_train[feature_names[0]].astype(np.int64)
X_test[feature_names[0]] = X_test[feature_names[0]].astype(np.int64)
model_task = ModelTask.CLASSIFICATION
matrix_features = [feature_names[0]]
run_error_analyzer_on_models(X_train,
y_train,
X_test,
y_test,
feature_names,
model_task,
matrix_features=matrix_features)
def test_matrix_filter_iris_filters(self):
# Validate the shift cohort functionality where base
# cohort was chosen in tree view
X_train, X_test, y_train, y_test, feature_names, _ = create_iris_data()
filters = [{
'arg': [2.85],
'column': feature_names[1],
'method': 'less and equal'
}]
model_task = ModelTask.CLASSIFICATION
run_error_analyzer_on_models(X_train,
y_train,
X_test,
y_test,
feature_names,
model_task,
filters=filters)
def test_explain_model_serialization_multiclass(self, mimic_explainer):
x_train, x_test, y_train, _, _, _ = create_iris_data()
# Fit an SVM model
model = create_sklearn_svm_classifier(x_train, y_train)
self._validate_model_serialization(model, x_train, x_test,
mimic_explainer)
