def test_safe_sparse_dot_2d_1d(A_array_constr):
rng = np.random.RandomState(0)
B = rng.random_sample((10))
# 2D @ 1D
A = rng.random_sample((30, 10))
expected = np.dot(A, B)
A = A_array_constr(A)
actual = safe_sparse_dot(A, B)
assert_allclose(actual, expected)
# 1D @ 2D
A = rng.random_sample((10, 30))
expected = np.dot(B, A)
A = A_array_constr(A)
actual = safe_sparse_dot(B, A)
assert_allclose(actual, expected)
def test_safe_sparse_dot_nd():
rng = np.random.RandomState(0)
# dense ND / sparse
A = rng.random_sample((2, 3, 4, 5, 6))
B = rng.random_sample((6, 7))
expected = np.dot(A, B)
B = sparse.csr_matrix(B)
actual = safe_sparse_dot(A, B)
assert_allclose(actual, expected)
# sparse / dense ND
A = rng.random_sample((2, 3))
B = rng.random_sample((4, 5, 3, 6))
expected = np.dot(A, B)
A = sparse.csr_matrix(A)
actual = safe_sparse_dot(A, B)
assert_allclose(actual, expected)
def test_safe_sparse_dot_2d(A_array_constr, B_array_constr):
rng = np.random.RandomState(0)
A = rng.random_sample((30, 10))
B = rng.random_sample((10, 20))
expected = np.dot(A, B)
A = A_array_constr(A)
B = B_array_constr(B)
actual = safe_sparse_dot(A, B, dense_output=True)
assert_allclose(actual, expected)
def test_safe_sparse_dot_dense_output(dense_output):
rng = np.random.RandomState(0)
A = sparse.random(30, 10, density=0.1, random_state=rng)
B = sparse.random(10, 20, density=0.1, random_state=rng)
expected = A.dot(B)
actual = safe_sparse_dot(A, B, dense_output=dense_output)
assert sparse.issparse(actual) == (not dense_output)
if dense_output:
expected = expected.toarray()
assert_allclose_dense_sparse(actual, expected)
def test_sparse_decision_function():
# Test decision_function
# Sanity check, test that decision_function implemented in python
# returns the same as the one in libsvm
# multi class:
svc = svm.SVC(kernel='linear', C=0.1, decision_function_shape='ovo')
clf = svc.fit(iris.data, iris.target)
dec = safe_sparse_dot(iris.data, clf.coef_.T) + clf.intercept_
assert_array_almost_equal(dec, clf.decision_function(iris.data))
# binary:
clf.fit(X, Y)
dec = np.dot(X, clf.coef_.T) + clf.intercept_
prediction = clf.predict(X)
assert_array_almost_equal(dec.ravel(), clf.decision_function(X))
assert_array_almost_equal(
prediction,
clf.classes_[(clf.decision_function(X) > 0).astype(np.int).ravel()])
expected = np.array([-1., -0.66, -1., 0.66, 1., 1.])
assert_array_almost_equal(clf.decision_function(X), expected, 2)
def kfunc(x, y):
return safe_sparse_dot(x, y.T)