def test_gp_search():
# Test that the best estimator contains the right value for foo_param
clf = MockDiscreteClassifier()
gp_search = GPSearchCV(clf, {'foo_param': ['int', [1, 3]]})
# make sure it selects the smallest parameter in case of ties
old_stdout = sys.stdout
sys.stdout = StringIO()
gp_search.fit(X, y)
sys.stdout = old_stdout
assert_equal(gp_search.best_estimator_.foo_param, 2)
clf = MockContinuousClassifier()
gp_search = GPSearchCV(clf, {'foo_param': ['float', [-3, 3]]}, verbose=3)
# make sure it selects the smallest parameter in case of ties
old_stdout = sys.stdout
sys.stdout = StringIO()
gp_search.fit(X, y)
sys.stdout = old_stdout
assert_almost_equal(gp_search.best_estimator_.foo_param, 0, decimal=1)
# Smoke test the score etc:
gp_search.score(X, y)
gp_search.predict_proba(X)
gp_search.decision_function(X)
gp_search.transform(X)
# Test exception handling on scoring
gp_search.scoring = 'sklearn'
assert_raises(ValueError, gp_search.fit, X, y)
def test_gp_search():
clf = MockClassifier()
gp_search = GPSearchCV(clf, {'foo_param': ['int', [1, 3]]}, verbose=3)
# make sure it selects the smallest parameter in case of ties
old_stdout = sys.stdout
sys.stdout = StringIO()
gp_search.fit(X, y)
sys.stdout = old_stdout
assert_equal(gp_search.best_estimator_.foo_param, 2)
for i, foo_i in enumerate([1, 2, 3]):
assert_true(gp_search.scores_[i][0]
== {'foo_param': foo_i})
# Smoke test the score etc:
gp_search.score(X, y)
gp_search.predict_proba(X)
gp_search.decision_function(X)
gp_search.transform(X)
# Test exception handling on scoring
gp_search.scoring = 'sklearn'
assert_raises(ValueError, gp_search.fit, X, y)
def test_grid_search_no_score():
# Test grid-search on classifier that has no score function.
clf = LinearSVC(random_state=0)
X, y = make_blobs(random_state=0, centers=2)
Cs = [.1, 1, 10]
clf_no_score = LinearSVCNoScore(random_state=0)
gp_search = GPSearchCV(clf, {'C': Cs}, scoring='accuracy')
gp_search.fit(X, y)
grid_search_no_score = GPSearchCV(clf_no_score, {'C': Cs},
scoring='accuracy')
# smoketest grid search
grid_search_no_score.fit(X, y)
# check that best params are equal
assert_equal(grid_search_no_score.best_params_, gp_search.best_params_)
# check that we can call score and that it gives the correct result
assert_equal(gp_search.score(X, y), grid_search_no_score.score(X, y))
# giving no scoring function raises an error
grid_search_no_score = GPSearchCV(clf_no_score, {'C': Cs})
assert_raise_message(TypeError, "no scoring", grid_search_no_score.fit,
[[1]])
