def test_fit():
X = X_dsel_ex1
y = y_dsel_ex1
pool_classifiers = create_pool_classifiers()
single_best_test = SingleBest(pool_classifiers)
single_best_test.fit(X, y)
assert single_best_test.best_clf_index == 0 or single_best_test.best_clf_index == 2
def test_not_predict_proba():
X = X_dsel_ex1
y = y_dsel_ex1
classifier = MagicMock()
classifier.predict.return_value = [0]
single_best_test = SingleBest([classifier] * 10)
single_best_test.fit(X, y)
with pytest.raises(ValueError):
single_best_test.predict_proba(X)
def test_predict():
X = X_dsel_ex1
y = y_dsel_ex1
pool_classifiers = create_pool_classifiers()
single_best_test = SingleBest(pool_classifiers)
single_best_test.fit(X, y)
predicted_labels = single_best_test.predict(X)
assert np.equal(predicted_labels, 0).all()
def test_predict_proba():
X = X_dsel_ex1
y = y_dsel_ex1
pool_classifiers = create_pool_classifiers()
single_best_test = SingleBest(pool_classifiers)
single_best_test.fit(X, y)
predicted_proba = single_best_test.predict_proba(X)
assert np.equal(predicted_proba, pool_classifiers[0].predict_proba(X)).all()
def test_predict(create_X_y, create_pool_classifiers):
X, y = create_X_y
pool_classifiers = create_pool_classifiers
single_best_test = SingleBest(pool_classifiers=pool_classifiers)
single_best_test.fit(X, y)
predicted_labels = single_best_test.predict(X)
assert np.equal(predicted_labels, 0).all()
def test_not_predict_proba(create_X_y):
X, y = create_X_y
classifier = MagicMock()
classifier.predict.return_value = [0]
single_best_test = SingleBest([classifier] * 10)
single_best_test.fit(X, y)
with pytest.raises(ValueError):
single_best_test.predict_proba(X)
def test_fit(create_X_y, create_pool_classifiers):
X, y = create_X_y
pool_classifiers = create_pool_classifiers
single_best_test = SingleBest(pool_classifiers)
single_best_test.fit(X, y)
assert (single_best_test.best_clf_index_ == 0
or single_best_test.best_clf_index_ == 2)
def test_predict_proba(create_X_y, create_pool_classifiers):
X, y = create_X_y
pool_classifiers = create_pool_classifiers
single_best_test = SingleBest(pool_classifiers)
single_best_test.fit(X, y)
predicted_proba = single_best_test.predict_proba(X)
assert np.equal(predicted_proba,
pool_classifiers[0].predict_proba(X)).all()
def test_label_encoder_base_ensemble():
from sklearn.ensemble import RandomForestClassifier
X, y = make_classification()
y[y == 1] = 2
y = y.astype(np.float)
pool = RandomForestClassifier().fit(X, y)
sb = SingleBest(pool)
sb.fit(X, y)
pred = sb.predict(X)
assert np.isin(sb.classes_, pred).all()
def test_fit(create_X_y, create_pool_classifiers):
X, y = create_X_y
pool_classifiers = create_pool_classifiers
single_best_test = SingleBest(pool_classifiers)
single_best_test._estimate_performances = MagicMock(
return_value=[1.0, 0.5, 0.99])
single_best_test.fit(X, y)
assert single_best_test.best_clf_index_ == 0
def test_different_scorer():
X, y = make_classification(n_samples=100, random_state=42)
X_val, y_val = make_classification(n_samples=25, random_state=123)
pool = AdaBoostClassifier(n_estimators=10).fit(X, y)
performances = []
for clf in pool:
preds = clf.predict_proba(X_val)
performances.append(roc_auc_score(y_val.ravel(), preds[:, -1]))
id_best = np.argmax(performances)
sb = SingleBest(pool_classifiers=pool, scoring='roc_auc')
sb.fit(X_val, y_val)
assert id_best == sb.best_clf_index_
def test_single_best():
pool_classifiers, X_dsel, y_dsel, X_test, y_test = setup_classifiers()
single_best = SingleBest(pool_classifiers)
single_best.fit(X_dsel, y_dsel)
assert np.isclose(single_best.score(X_test, y_test), 0.86969696969696975)