def test_scale_paretto(random_df):
data = random_df
scaled = utils.scale(data, "pareto")
assert np.isclose(scaled.mean(), 0).all()
# in pareto scaling the std of x is equal to the variance of x scaled
assert np.isclose(scaled.var(), data.std()).all()
def test_scale_autoscaling(random_df):
data = random_df
scaled = utils.scale(data, "autoscaling")
assert np.isclose(scaled.mean(), 0).all()
assert np.isclose(scaled.var(), 1).all()
def test_scale_rescaling():
data = np.random.normal(size=(200, 50), loc=10)
data = pd.DataFrame(data)
scaled = utils.scale(data, "rescaling")
assert np.isclose(scaled.min(), 0).all()
assert np.isclose(scaled.max(), 1).all()
def test_scale_invalid_mode(random_df):
data = random_df
with pytest.raises(ValueError):
utils.scale(data, "invalid_mode")