def test_p_normalize_zero_row(): def almost(a, b, prec=0.00001): return np.abs(a - b) < prec # no axis given X = np.array([ [0, 0, 0], [0, 0, 0] ]) Y = fl.p_normalize(X) assert Y.shape == (2, 3) assert almost(np.sum(Y), 1.0) X = np.array([ [0.0, 0.0, 0.0], [0, 0, 0] ]) Y = fl.p_normalize(X, 1) assert Y.shape == (2, 3) assert almost(np.sum(Y), 2.0) X = np.array([ [0.0, 0.0, 0.0], [0, 0, 0], ]) Y = fl.p_normalize(X, 0) assert Y.shape == (2, 3) assert almost(np.sum(Y), 3.0)
def test_p_normalize_zero_row(): # no axis given X = np.array([ [0, 0, 0], [0, 0, 0] ]) Y = fl.p_normalize(X) assert Y.shape == (2, 3) assert np.sum(Y) == 0.0 X = np.array([ [0.0, 0.0, 0.0], [0, 0, 0] ]) Y = fl.p_normalize(X, 1) assert Y.shape == (2, 3) assert np.sum(Y) == 0.0 X = np.array([ [0.0, 0.0, 0.0], [0, 0, 0], ]) Y = fl.p_normalize(X, 0) assert Y.shape == (2, 3) assert np.sum(Y) == 0.0
def test_p_normalize_zero_row(): # no axis given X = np.array([[0, 0, 0], [0, 0, 0]]) Y = fl.p_normalize(X) assert Y.shape == (2, 3) assert np.sum(Y) == 0.0 X = np.array([[0.0, 0.0, 0.0], [0, 0, 0]]) Y = fl.p_normalize(X, 1) assert Y.shape == (2, 3) assert np.sum(Y) == 0.0 X = np.array([ [0.0, 0.0, 0.0], [0, 0, 0], ]) Y = fl.p_normalize(X, 0) assert Y.shape == (2, 3) assert np.sum(Y) == 0.0
def test_p_normalize_wrong_dimensions(): with pytest.raises(AssertionError): X = np.array([[1, 2, 3]]) Y = fl.p_normalize(X, 2) with pytest.raises(AssertionError): X = np.array([[1, 2, 3]]) Y = fl.p_normalize(X, -1)
def test_p_normalize_wrong_dimensions(): with pytest.raises(ValueError): X = np.array([[1, 2, 3]]) Y = fl.p_normalize(X, 2) with pytest.raises(ValueError): X = np.array([[1, 2, 3]]) Y = fl.p_normalize(X, -1)
def test_p_normalize(): def almost(a, b): return np.abs(a - b) < 0.00001 X = np.array([1, 2, 3, 4]) Y = fl.p_normalize(X) assert almost(1. / 10, Y[0]) assert almost(2. / 10, Y[1]) assert almost(3. / 10, Y[2]) assert almost(4. / 10, Y[3]) X = np.array([ [1, 2, 3, 4], [4, 4, 2, 0], ]) Y = fl.p_normalize(X, 1) assert almost(1. / 10, Y[0][0]) assert almost(2. / 10, Y[0][1]) assert almost(3. / 10, Y[0][2]) assert almost(4. / 10, Y[0][3]) assert almost(4. / 10, Y[1][0]) assert almost(4. / 10, Y[1][1]) assert almost(2. / 10, Y[1][2]) assert almost(0. / 10, Y[1][3]) X = np.array([ [3, 9], [3, 1], ]) Y = fl.p_normalize(X, 0) assert 0.5 == Y[0][0] assert 0.5 == Y[1][0] assert 0.9 == Y[0][1] assert 0.1 == Y[1][1]