def test_quick_method(self):
"""
Test the quick method with a random dataset
"""
X, y = make_classification(
n_samples=400, n_features=20, n_informative=8, n_redundant=8,
n_classes=2, n_clusters_per_class=4, random_state=27
)
_, ax = plt.subplots()
classification_report(DecisionTreeClassifier(), X, y, ax=ax)
self.assert_images_similar(ax=ax)
def test_quick_method(self):
"""
Test the quick method with a random dataset
"""
X, y = make_classification(
n_samples=400, n_features=20, n_informative=8, n_redundant=8,
n_classes=2, n_clusters_per_class=4, random_state=27
)
_, ax = plt.subplots()
model = DecisionTreeClassifier(random_state=19)
classification_report(model, X, y, ax=ax, random_state=42)
self.assert_images_similar(ax=ax, tol=25.0)
def test_quick_method(self):
"""
Test the quick method with a random dataset
"""
X, y = make_classification(
n_samples=400,
n_features=20,
n_informative=8,
n_redundant=8,
n_classes=2,
n_clusters_per_class=4,
random_state=27,
)
# Create train/test splits
splits = tts(X, y, test_size=0.2, random_state=42)
X_train, X_test, y_train, y_test = splits
_, ax = plt.subplots()
model = DecisionTreeClassifier(random_state=19)
visualizer = classification_report(
model, X_train, y_train, X_test, y_test, ax=ax, show=False
)
assert isinstance(visualizer, ClassificationReport)
self.assert_images_similar(visualizer, tol=12)
def test_catboost_classifier(self):
"""
Validate CatBoost classifier with wrapper
"""
X, y = make_classification(n_samples=400,
n_features=10,
n_informative=2,
n_redundant=3,
n_classes=2,
n_clusters_per_class=2,
random_state=8311982)
X_train, X_test, y_train, y_test = tts(X, y)
model = classifier(
catboost.CatBoostClassifier(iterations=2,
depth=2,
learning_rate=1,
loss_function='Logloss'))
# For some reason, this works if you call fit directly and pass is_fitted to
# the visualizer, but does not work if you rely on the visualizer to fit the
# model on the data. I can't tell if this is a catboost or Yellowbrick issue.
model.fit(X_train, y_train)
oz = classification_report(model,
X_train,
y_train,
X_test,
y_test,
is_fitted=True,
show=False)
assert is_fitted(oz)
def test_cuml_classifier(self):
"""
Validate cuML classifier with wrapper
"""
# NOTE: this is currently untested as I wasn't able to install cuML
X, y = make_classification(n_samples=400,
n_features=10,
n_informative=2,
n_redundant=3,
n_classes=2,
n_clusters_per_class=2,
random_state=8311982)
X_train, X_test, y_train, y_test = tts(X, y)
# Convert to cudf dataframes
X_train = cudf.DataFrame(X_train)
y_train = cudf.Series(y_train)
X_test = cudf.DataFrame(X_test)
y_test = cudf.Series(y_test)
model = classifier(curfc(n_estimators=40, max_depth=8, max_features=1))
oz = classification_report(model,
X_train,
y_train,
X_test,
y_test,
show=False)
assert is_fitted(oz)
verbose_feature_names = lines
feature_names = np.arange(len(verbose_feature_names))
# %%
class_names = ['+', '-']
verbose_class_names = ['Activator', 'Repressor']
# %%
X = load('X.joblib')
y = load('y.joblib')
# %%
to_graphviz(clf, num_trees=0, rankdir='LR')
# %%
classification_report(clf, X, y)
# %%
visualizer = ROCAUC(clf, classes=class_names)
visualizer.score(X, y)
visualizer.poof()
# %%
visualizer = ClassPredictionError(clf, classes=class_names)
visualizer.score(X, y)
visualizer.poof()
# %%
visualizer = DiscriminationThreshold(clf)
visualizer.fit(X, y)
visualizer.poof()