def test__clf_with_no_proba_fail():
X, y = iris_data()
clf = OneRClassifier()
clf.fit(X, y)
x_ref = X[15]
s = ("Your `model` does not support "
"`predict_proba`. Set `y_desired_proba` "
" to `None` to use `predict`instead.")
assert_raises(AttributeError, s, create_counterfactual, x_ref, 2, clf, X,
1., 100, 123)
def test_iris_quartiles_resolve_ties_chi_squared_2():
# tests with duplicate column
oner = OneRClassifier(resolve_ties='chi-squared')
Xd_traintemp = np.zeros((Xd_train.shape[0], Xd_train.shape[1]+1))
Xd_traintemp[:, 0] = Xd_train[:, 2]
Xd_traintemp[:, 1] = Xd_train[:, 0]
Xd_traintemp[:, 2] = Xd_train[:, 1]
Xd_traintemp[:, 3] = Xd_train[:, 2]
Xd_traintemp[:, 4] = Xd_train[:, 3]
oner.fit(Xd_traintemp, y_train)
assert oner.feature_idx_ == 0
assert oner.prediction_dict_['total error'] == 16
assert round(oner.score(Xd_traintemp, y_train), 4) == 0.8571
def test_iris_quartiles_resolve_ties_first():
oner = OneRClassifier()
oner.fit(Xd_train, y_train)
assert oner.feature_idx_ == 2
assert oner.prediction_dict_['total error'] == 16
assert round(oner.score(Xd_train, y_train), 4) == 0.8571
assert round(oner.score(Xd_test, y_test), 4) == 0.8421
def test_iris_quartiles_resolve_ties_chi_squared_1():
oner = OneRClassifier(resolve_ties='chi-squared')
oner.fit(Xd_train, y_train)
assert oner.feature_idx_ == 2
assert oner.prediction_dict_['total error'] == 16
assert round(oner.score(Xd_train, y_train), 4) == 0.8571
assert round(oner.score(Xd_test, y_test), 4) == 0.8421
np.testing.assert_almost_equal(oner.p_value_, 0.0, decimal=7)
def test__clf_with_no_proba_pass():
X, y = iris_data()
clf = OneRClassifier()
clf.fit(X, y)
x_ref = X[15]
res = create_counterfactual(x_reference=x_ref,
y_desired=2,
model=clf,
X_dataset=X,
y_desired_proba=None,
lammbda=100,
random_seed=123)
assert clf.predict(x_ref.reshape(1, -1)) == 0
assert clf.predict(res.reshape(1, -1)) == 2