def test_basic():
y_true = np.array([False, True, True, True, False, False, False, True])
y_pred = np.array([0.491, -0.1, 0.64, 1.52, -0.23, -0.23, 1.579, 0.76])
phi = pearson(y_true, y_pred)
reference = 0.2225646085633681
assert abs(phi - reference) < ATOL
rho = spearman(y_true, y_pred)
reference = 0.5714285714285714
assert abs(rho - reference) < ATOL
def test_basic100():
rng = np.random.RandomState(42)
y_true = rng.randn(100)
y_pred = rng.randn(y_true.size)
phi = pearson(y_true, y_pred)
reference = -0.13642221217000247
assert abs(phi - reference) < ATOL
rho = spearman(y_true, y_pred)
reference = -0.10796279627962799
assert abs(rho - reference) < ATOL
def test_basic_vs_scipy():
rng = np.random.RandomState(42)
y_true = rng.randn(1000)
y_pred = rng.randn(y_true.size)
phi = pearson(y_true, y_pred)
reference = stats.pearsonr(y_true, y_pred)[0]
assert abs(phi - reference) < ATOL
rho = spearman(y_true, y_pred)
reference = stats.spearmanr(y_true, y_pred)[0]
assert abs(rho - reference) < ATOL