def test_remaining_obs_put_in_last_slice():
"""This was a bug found in testing. This particular dataset puts 9
counts"""
y = np.arange(100)
slice_indicator, counts = slice_y(y, n_slices=11)
np.testing.assert_array_equal(np.hstack((np.repeat(9, 10), 10)), counts)
expected_slices = np.hstack((np.repeat(np.arange(10),
9), np.repeat(10, 10)))
np.testing.assert_array_equal(expected_slices, slice_indicator)
def test_uneven_slices_within_y():
"""Make sure the slice function actually groups common y values together.
"""
y = np.hstack((np.repeat(1, 13), np.repeat(2, 5), np.repeat(3, 25)))
slice_indicator, counts = slice_y(y, n_slices=3)
np.testing.assert_array_equal(counts, np.array([13, 5, 25]))
expected_indicator = np.hstack(
(np.repeat(0, 13), np.repeat(1, 5), np.repeat(2, 25)))
np.testing.assert_array_equal(slice_indicator, expected_indicator)
def test_uneven_slices():
"""Too match DR the slice function spills over to 6 slices if it cannot
fit it into 5..."""
shuffle = module_rng.permutation(68)
y = np.arange(68)[shuffle]
slice_indicator, counts = slice_y(y, n_slices=5)
np.testing.assert_array_equal(counts, np.array([13, 13, 13, 13, 13, 3]))
np.testing.assert_array_equal(
slice_indicator,
np.hstack((np.repeat(np.arange(5), 13), np.array([5, 5, 5]))))
def test_slice_y_simple():
"""Test a simple a target with 100 unique values and even slices."""
shuffle = module_rng.permutation(100)
y = np.arange(100)[shuffle]
slice_indicator, counts = slice_y(y, n_slices=10)
# all slices should have 10 counts
np.testing.assert_array_equal(counts, np.repeat(10, 10))
# continguous indices counting up from 10
np.testing.assert_array_equal(slice_indicator,
np.repeat(np.arange(10), 10))
def test_slice_raises_error_when_y_has_one_value():
"""Error if y only has one value."""
y = np.ones(100)
with pytest.raises(ValueError):
slice_y(y)
def test_slice_match_athletes():
"""Test that we match the slices used in the DR package on the
athlete's dataset.
"""
X, y = load_athletes()
slice_indicator, counts = slice_y(y, n_slices=11)
# counts per slice match DR package
expected_counts = np.array([18, 18, 18, 18, 18, 19, 18, 19, 23, 18, 15])
np.testing.assert_array_equal(counts, expected_counts)
# slice indicators match DR package
expected_indicator = np.array([
6,
5,
4,
4,
2,
3,
5,
1,
3,
3,
8,
6,
1,
7,
5,
4,
3,
4,
6,
4,
6,
6,
4,
3,
7,
6,
5,
4,
1,
2,
2,
5,
5,
6,
6,
2,
1,
3,
2,
4,
4,
2,
3,
4,
2,
5,
4,
5,
3,
1,
3,
4,
3,
5,
5,
3,
3,
3,
4,
2,
5,
5,
6,
5,
4,
2,
2,
2,
5,
8,
6,
6,
7,
5,
8,
1,
3,
1,
3,
1,
3,
1,
2,
3,
2,
1,
2,
2,
2,
4,
4,
1,
2,
1,
2,
1,
1,
1,
1,
1,
5,
8,
9,
10,
9,
8,
8,
9,
9,
10,
7,
9,
11,
9,
9,
9,
11,
9,
10,
10,
1,
10,
10,
10,
10,
11,
10,
9,
11,
8,
10,
10,
11,
11,
11,
9,
10,
8,
9,
8,
4,
7,
7,
8,
11,
10,
9,
6,
7,
8,
6,
5,
4,
8,
3,
7,
7,
9,
9,
11,
9,
9,
11,
7,
9,
8,
7,
6,
6,
6,
7,
7,
6,
6,
5,
11,
11,
11,
10,
7,
8,
9,
7,
9,
8,
8,
9,
7,
11,
9,
9,
11,
8,
10,
10,
10,
7,
10,
8,
7,
6,
8,
])
np.testing.assert_array_equal(slice_indicator,
np.sort(expected_indicator) - 1)