def test_input_dimensions():
with pytest.raises(ValueError) as error:
M = np.array([0, -1, 1])
check_rating_matrix(M)
assert 'got 1 dimension(s)' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.zeros((1, 2, 3))
check_rating_matrix(M)
assert 'got 3 dimension(s)' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([0, -1, 1])
calibrate(M)
assert 'got 1 dimension(s)' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([0, -1, 1])
average(M)
assert 'got 1 dimension(s)' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([0, -1, 1])
find_min(M)
assert 'got 1 dimension(s)' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([0, -1, 1])
find_max(M)
assert 'got 1 dimension(s)' in str(error.value)
def test_performance_review_0():
M = np.loadtxt('tests/rat_mat_0.data')
assert M.shape[0] == M.shape[1]
n = M.shape[0]
ratings, distinct, lenient = calibrate(M, (0, 100))
assert len(ratings) == n and len(distinct) == n and len(lenient) == n
assert np.isclose(ratings.min(), 0) and np.isclose(ratings.max(), 100)
def test_performance_review_1():
M = np.loadtxt('tests/rat_mat_1.data')
assert M.shape[0] == M.shape[1]
n = M.shape[0]
avg_rat = average(M)
ratings, distinct, lenient = calibrate(M, (avg_rat.min(), avg_rat.max()))
assert len(ratings) == n and len(distinct) == n and len(lenient) == n
assert np.isclose(ratings.min(), avg_rat.min()) and np.isclose(
ratings.max(), avg_rat.max())
def test_input_ratings():
with pytest.raises(ValueError) as error:
M = np.array([[2, 1, 0.5], [5, -2, 1]])
calibrate(M)
assert 'should be -1 denoting null' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([[2, 1, 0.5], [5, -2, 1]])
average(M)
assert 'should be -1 denoting null' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([[2, 1, 0.5], [5, -2, 1]])
find_min(M)
assert 'should be -1 denoting null' in str(error.value)
with pytest.raises(ValueError) as error:
M = np.array([[2, 1, 0.5], [5, -2, 1]])
find_max(M)
assert 'should be -1 denoting null' in str(error.value)
def test_paper_review_example():
M = np.array([[.40, .44, -1., .67, -1., .81, .89, .95, .99],
[.15, .19, .32, .45, .49, .62, -1., -1., -1.],
[-1., .21, -1., .39, -1., .61, -1., .79, -1.],
[.20, -1., .20, -1., .60, -1., .70, -1., .80],
[-1., -1., .10, -1., .20, -1., .30, .40, .50]])
ratings, distinct, lenient = calibrate(M, (0.0, 1.0), False)
assert np.array_equal(np.round(ratings, decimals=2),
[0.0, .05, .18, .40, .53, .68, .75, .92, 1.0])
avg_rat = average(M)
assert np.array_equal(np.round(avg_rat, decimals=2),
[.25, .28, .21, .50, .43, .68, .63, .71, .76])
assert find_min(M) == .10
assert find_max(M) == .99