Skip to content
Snippets Groups Projects
Select Git revision
  • main default protected
  • robertos-branch
  • aifmd_dataset_support
  • v1.3.1
  • v1.3.0
  • v1.2.0
  • v1.1.0
  • v1.0.2
  • v1.0.1
  • v1.0.0
  • v1.0.0b0
  • v1.0.0a0
12 results
Blame Permalink
test_shenzhen.py 3.00 KiB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111 






# SPDX-FileCopyrightText: Copyright © 2023 Idiap Research Institute <contact@idiap.ch>
#
# SPDX-License-Identifier: GPL-3.0-or-later
"""Tests for Shenzhen dataset."""

import importlib

import pytest


def id_function(val):
    if isinstance(val, dict):
        return str(val)
    return repr(val)


@pytest.mark.parametrize(
    "split,lenghts",
    [
        ("default", dict(train=422, validation=107, test=133)),
        ("fold-0", dict(train=476, validation=119, test=67)),
        ("fold-1", dict(train=476, validation=119, test=67)),
        ("fold-2", dict(train=476, validation=120, test=66)),
        ("fold-3", dict(train=476, validation=120, test=66)),
        ("fold-4", dict(train=476, validation=120, test=66)),
        ("fold-5", dict(train=476, validation=120, test=66)),
        ("fold-6", dict(train=476, validation=120, test=66)),
        ("fold-7", dict(train=476, validation=120, test=66)),
        ("fold-8", dict(train=476, validation=120, test=66)),
        ("fold-9", dict(train=476, validation=120, test=66)),
    ],
    ids=id_function,  # just changes how pytest prints it
)
def test_protocol_consistency(
    database_checkers, split: str, lenghts: dict[str, int]
):
    from mednet.config.data.shenzhen.datamodule import make_split

    database_checkers.check_split(
        make_split(f"{split}.json"),
        lengths=lenghts,
        prefixes=("CXR_png/CHNCXR_0",),
        possible_labels=(0, 1),
    )


@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
@pytest.mark.parametrize(
    "dataset",
    [
        "train",
        "validation",
        "test",
    ],
)
@pytest.mark.parametrize(
    "name",
    [
        "default",
        "fold_0",
        "fold_1",
        "fold_2",
        "fold_3",
        "fold_4",
        "fold_5",
        "fold_6",
        "fold_7",
        "fold_8",
        "fold_9",
    ],
)
def test_loading(database_checkers, name: str, dataset: str):
    datamodule = importlib.import_module(
        f".{name}", "mednet.config.data.shenzhen"
    ).datamodule

    datamodule.model_transforms = []  # should be done before setup()
    datamodule.setup("predict")  # sets up all datasets

    loader = datamodule.predict_dataloader()[dataset]

    limit = 3  # limit load checking
    for batch in loader:
        if limit == 0:
            break
        database_checkers.check_loaded_batch(
            batch,
            batch_size=1,
            color_planes=1,
            prefixes=("CXR_png/CHNCXR_0",),
            possible_labels=(0, 1),
            expected_num_labels=1,
        )
        limit -= 1


@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
def test_loaded_image_quality(database_checkers, datadir):
    reference_histogram_file = str(
        datadir / "histograms/raw_data/histograms_shenzhen_default.json"
    )

    datamodule = importlib.import_module(
        ".default", "mednet.config.data.shenzhen"
    ).datamodule

    datamodule.model_transforms = []
    datamodule.setup("predict")

    database_checkers.check_image_quality(datamodule, reference_histogram_file)