def test_pcs_efficient_false_dense_output_false():
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=False,
dense_output=False)
assert isspmatrix(output), "The output is not sparse."
assert output.sum() == approx(8.202465881449182), "Output not correct."
def test_pcs_efficient_true_multi_dimensional():
output_1 = pairwise_cosine_similarity(A=E,
B=F,
efficient=True,
n_top=3,
lower_bound=0.1,
n_jobs=-1,
dense_output=True)
output_2 = pairwise_cosine_similarity(A=G,
B=H,
efficient=True,
n_top=3,
lower_bound=0.1,
n_jobs=-1,
dense_output=True)
assert output_1.sum() == approx(output_2.sum()), "Multi-dimensional error."
def test_pcs_efficient_true_n_jobs():
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=True,
n_top=4,
lower_bound=0.0,
n_jobs=-1,
dense_output=True)
assert output.sum() == approx(8.202465881449182), "Output not correct."
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=True,
n_top=4,
lower_bound=0.0,
n_jobs=1,
dense_output=True)
assert output.sum() == approx(8.202465881449182), "Output not correct."
def test_pcs_efficient_true_dense_output_true():
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=True,
n_top=4,
lower_bound=0.0,
n_jobs=2,
dense_output=True)
assert not isspmatrix(output), "The output is sparse."
assert output.sum() == approx(8.202465881449182), "Output not correct."
def test_pcs_efficient_true_sparse():
output = pairwise_cosine_similarity(A=C,
B=D,
efficient=True,
n_top=2,
lower_bound=0.1,
n_jobs=-1,
dense_output=True)
assert output.sum() == approx(126.79083764959829), "Sparse input error."
def test_pcs_efficient_true_n_top():
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=True,
n_top=2,
lower_bound=0.1,
n_jobs=-1,
dense_output=True)
assert (output > 0).sum() == 6, "Error in n_top."
def test_pcs_efficient_true_lower_bound():
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=True,
n_top=4,
lower_bound=0.9,
n_jobs=-1,
dense_output=True)
assert (output > 0).sum() == 5, "Error in lower_bound."
output = pairwise_cosine_similarity(A=A,
B=B,
efficient=True,
n_top=4,
lower_bound=0.99,
n_jobs=-1,
dense_output=True)
assert (output > 0).sum() == 1, "Error in lower_bound."