Python to_onehot Exemples

Exemple #1

Fichier : test_classification.py Projet : yoshum/pytorch-lightning

def test_f1_score(pred, target, exp_score):
    score = f1_score(torch.tensor(pred), torch.tensor(target), class_reduction='none')
    assert torch.allclose(score, torch.tensor(exp_score))

    score = f1_score(to_onehot(torch.tensor(pred)), torch.tensor(target), class_reduction='none')
    assert torch.allclose(score, torch.tensor(exp_score))

Exemple #2

        pytest.param(torch.rand(32, 10, 28, 28), torch.randint(
            10, (32, 28, 28)), None, 10),
        pytest.param(torch.rand(32, 28, 28), torch.randint(10, (32, 28, 28)),
                     None, 10),
    ])
def test_get_num_classes(pred, target, num_classes, expected_num_classes):
    assert get_num_classes(pred, target, num_classes) == expected_num_classes


@pytest.mark.parametrize([
    'pred', 'target', 'expected_tp', 'expected_fp', 'expected_tn',
    'expected_fn', 'expected_support'
], [
    pytest.param(torch.tensor([0., 2., 4., 4.]), torch.tensor(
        [0., 4., 3., 4.]), 1, 1, 1, 1, 2),
    pytest.param(to_onehot(torch.tensor([0., 2., 4., 4.])),
                 torch.tensor([0., 4., 3., 4.]), 1, 1, 1, 1, 2)
])
def test_stat_scores(pred, target, expected_tp, expected_fp, expected_tn,
                     expected_fn, expected_support):
    tp, fp, tn, fn, sup = stat_scores(pred, target, class_index=4)

    assert tp.item() == expected_tp
    assert fp.item() == expected_fp
    assert tn.item() == expected_tn
    assert fn.item() == expected_fn
    assert sup.item() == expected_support


@pytest.mark.parametrize([
    'pred', 'target', 'expected_tp', 'expected_fp', 'expected_tn',

Exemple #3

Fichier : test_classification.py Projet : yoshum/pytorch-lightning

    assert torch.allclose(result, expected.to(result.dtype))


@pytest.mark.parametrize(['pred', 'target', 'num_classes', 'expected_num_classes'], [
    pytest.param(torch.rand(32, 10, 28, 28), torch.randint(10, (32, 28, 28)), 10, 10),
    pytest.param(torch.rand(32, 10, 28, 28), torch.randint(10, (32, 28, 28)), None, 10),
    pytest.param(torch.rand(32, 28, 28), torch.randint(10, (32, 28, 28)), None, 10),
])
def test_get_num_classes(pred, target, num_classes, expected_num_classes):
    assert get_num_classes(pred, target, num_classes) == expected_num_classes


@pytest.mark.parametrize(['pred', 'target', 'expected_tp', 'expected_fp',
                          'expected_tn', 'expected_fn', 'expected_support'], [
    pytest.param(torch.tensor([0., 2., 4., 4.]), torch.tensor([0., 4., 3., 4.]), 1, 1, 1, 1, 2),
    pytest.param(to_onehot(torch.tensor([0., 2., 4., 4.])), torch.tensor([0., 4., 3., 4.]), 1, 1, 1, 1, 2)
])
def test_stat_scores(pred, target, expected_tp, expected_fp, expected_tn, expected_fn, expected_support):
    tp, fp, tn, fn, sup = stat_scores(pred, target, class_index=4)

    assert tp.item() == expected_tp
    assert fp.item() == expected_fp
    assert tn.item() == expected_tn
    assert fn.item() == expected_fn
    assert sup.item() == expected_support


@pytest.mark.parametrize(['pred', 'target', 'reduction', 'expected_tp', 'expected_fp',
                          'expected_tn', 'expected_fn', 'expected_support'], [
    pytest.param(torch.tensor([0., 2., 4., 4.]), torch.tensor([0., 4., 3., 4.]), 'none',
                 [1, 0, 0, 0, 1], [0, 0, 1, 0, 1], [3, 4, 3, 3, 1], [0, 0, 0, 1, 1], [1, 0, 0, 1, 2]),

Exemple #4

Fichier : test_classification.py Projet : rhythmswing/pytorch-lightning

    assert torch.allclose(result, expected.to(result.dtype))


@pytest.mark.parametrize(['pred', 'target', 'num_classes', 'expected_num_classes'], [
    pytest.param(torch.rand(32, 10, 28, 28), torch.randint(10, (32, 28, 28)), 10, 10),
    pytest.param(torch.rand(32, 10, 28, 28), torch.randint(10, (32, 28, 28)), None, 10),
    pytest.param(torch.rand(32, 28, 28), torch.randint(10, (32, 28, 28)), None, 10),
])
def test_get_num_classes(pred, target, num_classes, expected_num_classes):
    assert get_num_classes(pred, target, num_classes) == expected_num_classes


@pytest.mark.parametrize(['pred', 'target', 'expected_tp', 'expected_fp',
                          'expected_tn', 'expected_fn', 'expected_support'], [
    pytest.param(torch.tensor([0., 2., 4., 4.]), torch.tensor([0., 4., 3., 4.]), 1, 1, 1, 1, 2),
    pytest.param(to_onehot(torch.tensor([0., 2., 4., 4.])), torch.tensor([0., 4., 3., 4.]), 1, 1, 1, 1, 2)
])
def test_stat_scores(pred, target, expected_tp, expected_fp, expected_tn, expected_fn, expected_support):
    tp, fp, tn, fn, sup = stat_scores(pred, target, class_index=4)

    assert tp.item() == expected_tp
    assert fp.item() == expected_fp
    assert tn.item() == expected_tn
    assert fn.item() == expected_fn
    assert sup.item() == expected_support


@pytest.mark.parametrize(['pred', 'target', 'expected_tp', 'expected_fp',
                          'expected_tn', 'expected_fn', 'expected_support'], [
    pytest.param(torch.tensor([0., 2., 4., 4.]), torch.tensor([0., 4., 3., 4.]),
                 [1, 0, 0, 0, 1], [0, 0, 1, 0, 1], [3, 4, 3, 3, 1], [0, 0, 0, 1, 1], [1, 0, 0, 1, 2]),