mindspore2022/tests/ut/python/dataset/test_datasets_stl10.py

# Copyright 2021 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""
Test STL10 dataset operators
"""
import os

import matplotlib.pyplot as plt
import numpy as np
import pytest

import mindspore.dataset as ds
import mindspore.dataset.vision.c_transforms as vision
from mindspore import log as logger

DATA_DIR = "../data/dataset/testSTL10Data"
WRONG_DIR = "../data/dataset/testMnistData"


def loadfile(path_to_data, path_to_labels=None):
    """
    Feature: loadfile.
    Description: parse stl10 file.
    Expectation: get image and label of stl10 dataset.
    """
    labels = None
    if path_to_labels:
        with open(os.path.realpath(path_to_labels), 'rb') as f:
            labels = np.fromfile(f, dtype=np.uint8) - 1  # 0-based

    with open(path_to_data, 'rb') as f:
        # read whole file in uint8 chunks
        everything = np.fromfile(f, dtype=np.uint8)
        images = np.reshape(everything, (-1, 3, 96, 96))
        images = np.transpose(images, (0, 1, 3, 2))
    return images, labels


def load_stl10(path, usage):
    """
    Feature: load_stl10.
    Description: load stl10.
    Expectation: get data of stl10 dataset.
    """
    assert usage in ["train", "test", "unlabeled", "train+unlabeled", "all"]

    if usage == "train":
        image_path = os.path.join(path, "train_X.bin")
        label_path = os.path.join(path, "train_y.bin")
        images, labels = loadfile(image_path, label_path)

    elif usage == "train+unlabeled":
        image_path = os.path.join(path, "train_X.bin")
        label_path = os.path.join(path, "train_y.bin")
        images, labels = loadfile(image_path, label_path)

        image_path = os.path.join(path, "unlabeled_X.bin")
        unlabeled_image, _ = loadfile(image_path)

        images = np.concatenate((images, unlabeled_image))
        labels = np.concatenate((labels, np.asarray([-1] * unlabeled_image.shape[0])))

    elif usage == "unlabeled":
        image_path = os.path.join(path, "unlabeled_X.bin")

        images, _ = loadfile(image_path)
        labels = np.asarray([-1] * images.shape[0])

    elif usage == "test":
        image_path = os.path.join(path, "test_X.bin")
        label_path = os.path.join(path, "test_y.bin")

        images, labels = loadfile(image_path, label_path)

    elif usage == "all":
        image_path = os.path.join(path, "test_X.bin")
        label_path = os.path.join(path, "test_y.bin")
        images, labels = loadfile(image_path, label_path)

        image_path = os.path.join(path, "train_X.bin")
        label_path = os.path.join(path, "train_y.bin")

        train_image, train_label = loadfile(image_path, label_path)

        images = np.concatenate((images, train_image))
        labels = np.concatenate((labels, train_label))

        image_path = os.path.join(path, "unlabeled_X.bin")
        unlabeled_image, _ = loadfile(image_path)

        images = np.concatenate((images, unlabeled_image))
        labels = np.concatenate((labels, np.asarray([-1] * unlabeled_image.shape[0])))

    return images, labels


def visualize_dataset(images, labels):
    """
    Feature: visualize_dataset.
    Description: visualize stl10 dataset.
    Expectation: plot images.
    """
    num_samples = len(images)
    for i in range(num_samples):
        plt.subplot(1, num_samples, i + 1)
        plt.imshow(np.transpose(images[i], (1, 2, 0)))
        plt.title(labels[i])
    plt.show()


def test_stl10_content_check():
    """
    Feature: test_stl10_content_check.
    Description: validate STL10ataset image readings.
    Expectation: get correct number of data and correct content.
    """
    logger.info("Test STL10Dataset Op with content check")
    # 1. train data.
    data1 = ds.STL10Dataset(DATA_DIR, usage="train", num_samples=1, shuffle=False)
    images, labels = load_stl10(DATA_DIR, "train")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label".
    for i, d in enumerate(data1.create_dict_iterator(num_epochs=1, output_numpy=True)):
        np.testing.assert_array_equal(d["image"], np.transpose(images[i], (1, 2, 0)))
        np.testing.assert_array_equal(d["label"], labels[i])
        num_iter += 1
    assert num_iter == 1

    # 2. test data.
    data1 = ds.STL10Dataset(DATA_DIR, usage="test", num_samples=1, shuffle=False)
    images, labels = load_stl10(DATA_DIR, "test")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label".
    for i, d in enumerate(data1.create_dict_iterator(num_epochs=1, output_numpy=True)):
        np.testing.assert_array_equal(d["image"], np.transpose(images[i], (1, 2, 0)))
        np.testing.assert_array_equal(d["label"], labels[i])
        num_iter += 1
    assert num_iter == 1

    # 3. unlabeled data.
    data1 = ds.STL10Dataset(DATA_DIR, usage="unlabeled", num_samples=1, shuffle=False)
    images, labels = load_stl10(DATA_DIR, "unlabeled")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label".
    for i, d in enumerate(data1.create_dict_iterator(num_epochs=1, output_numpy=True)):
        np.testing.assert_array_equal(d["image"], np.transpose(images[i], (1, 2, 0)))
        np.testing.assert_array_equal(d["label"], labels[i])
        num_iter += 1
    assert num_iter == 1

    # 4. train+unlabeled data.
    data1 = ds.STL10Dataset(DATA_DIR, usage="train+unlabeled", num_samples=2, shuffle=False)
    images, labels = load_stl10(DATA_DIR, "train+unlabeled")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label".
    for i, d in enumerate(data1.create_dict_iterator(num_epochs=1, output_numpy=True)):
        np.testing.assert_array_equal(d["image"], np.transpose(images[i], (1, 2, 0)))
        np.testing.assert_array_equal(d["label"], labels[i])
        num_iter += 1
    assert num_iter == 2

    # 4. all data.
    data1 = ds.STL10Dataset(DATA_DIR, usage="all", num_samples=3, shuffle=False)
    images, labels = load_stl10(DATA_DIR, "all")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label".
    for i, d in enumerate(data1.create_dict_iterator(num_epochs=1, output_numpy=True)):
        np.testing.assert_array_equal(d["image"], np.transpose(images[i], (1, 2, 0)))
        np.testing.assert_array_equal(d["label"], labels[i])
        num_iter += 1
    assert num_iter == 3


def test_stl10_basic():
    """
    Feature: test_stl10_basic.
    Description: test basic usage of STL10Dataset.
    Expectation: get correct number of data.
    """
    logger.info("Test STL10Dataset Op")

    # case 1: test loading whole dataset.
    all_data = ds.STL10Dataset(DATA_DIR, "all")
    num_iter = 0
    for _ in all_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 3

    # case 2: test num_samples.
    all_data = ds.STL10Dataset(DATA_DIR, "all", num_samples=1)
    num_iter = 0
    for _ in all_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 1

    # case 3: test repeat.
    all_data = ds.STL10Dataset(DATA_DIR, "all", num_samples=2)
    all_data = all_data.repeat(5)
    num_iter = 0
    for _ in all_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 10

    # case 4: test batch with drop_remainder=False.
    all_data = ds.STL10Dataset(DATA_DIR, "all", num_samples=2)
    assert all_data.get_dataset_size() == 2
    assert all_data.get_batch_size() == 1
    all_data = all_data.batch(batch_size=2)  # drop_remainder is default to be False.
    assert all_data.get_batch_size() == 2
    assert all_data.get_dataset_size() == 1

    num_iter = 0
    for _ in all_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 1

    # case 5: test batch with drop_remainder=True.
    all_data = ds.STL10Dataset(DATA_DIR, "all", num_samples=2)
    assert all_data.get_dataset_size() == 2
    assert all_data.get_batch_size() == 1
    all_data = all_data.batch(batch_size=2, drop_remainder=True)  # the rest of incomplete batch will be dropped.
    assert all_data.get_dataset_size() == 1
    assert all_data.get_batch_size() == 2
    num_iter = 0
    for _ in all_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 1


def test_stl10_sequential_sampler():
    """
    Feature: test_stl10_sequential_sampler.
    Description: test usage of STL10Dataset with SequentialSampler.
    Expectation: get correct number of data.
    """
    logger.info("Test STL10Dataset Op with SequentialSampler")
    num_samples = 2
    sampler = ds.SequentialSampler(num_samples=num_samples)
    all_data_1 = ds.STL10Dataset(DATA_DIR, "all", sampler=sampler)
    all_data_2 = ds.STL10Dataset(DATA_DIR, "all", shuffle=False, num_samples=num_samples)
    label_list_1, label_list_2 = [], []
    num_iter = 0
    for item1, item2 in zip(all_data_1.create_dict_iterator(num_epochs=1),
                            all_data_2.create_dict_iterator(num_epochs=1)):
        label_list_1.append(item1["label"].asnumpy())
        label_list_2.append(item2["label"].asnumpy())
        num_iter += 1
    np.testing.assert_array_equal(label_list_1, label_list_2)
    assert num_iter == num_samples


def test_stl10_exception():
    """
    Feature: test_stl10_exception.
    Description: test error cases for STL10Dataset.
    Expectation: raise exception.
    """
    logger.info("Test error cases for STL10Dataset")
    error_msg_1 = "sampler and shuffle cannot be specified at the same time"
    with pytest.raises(RuntimeError, match=error_msg_1):
        ds.STL10Dataset(DATA_DIR, "all", shuffle=False, sampler=ds.PKSampler(3))

    error_msg_2 = "sampler and sharding cannot be specified at the same time"
    with pytest.raises(RuntimeError, match=error_msg_2):
        ds.STL10Dataset(DATA_DIR, "all", sampler=ds.PKSampler(3), num_shards=2, shard_id=0)

    error_msg_3 = "num_shards is specified and currently requires shard_id as well"
    with pytest.raises(RuntimeError, match=error_msg_3):
        ds.STL10Dataset(DATA_DIR, "all", num_shards=10)

    error_msg_4 = "shard_id is specified but num_shards is not"
    with pytest.raises(RuntimeError, match=error_msg_4):
        ds.STL10Dataset(DATA_DIR, "all", shard_id=0)

    error_msg_5 = "Input shard_id is not within the required interval"
    with pytest.raises(ValueError, match=error_msg_5):
        ds.STL10Dataset(DATA_DIR, "all", num_shards=5, shard_id=-1)
    with pytest.raises(ValueError, match=error_msg_5):
        ds.STL10Dataset(DATA_DIR, "all", num_shards=5, shard_id=5)
    with pytest.raises(ValueError, match=error_msg_5):
        ds.STL10Dataset(DATA_DIR, "all", num_shards=2, shard_id=5)
    error_msg_6 = "num_parallel_workers exceeds"
    with pytest.raises(ValueError, match=error_msg_6):
        ds.STL10Dataset(DATA_DIR, "all", shuffle=False, num_parallel_workers=0)
    with pytest.raises(ValueError, match=error_msg_6):
        ds.STL10Dataset(DATA_DIR, "all", shuffle=False, num_parallel_workers=256)
    with pytest.raises(ValueError, match=error_msg_6):
        ds.STL10Dataset(DATA_DIR, "all", shuffle=False, num_parallel_workers=-2)
    error_msg_7 = "Argument shard_id"
    with pytest.raises(TypeError, match=error_msg_7):
        ds.STL10Dataset(DATA_DIR, "all", num_shards=2, shard_id="0")

    def exception_func(item):
        raise Exception("Error occur!")

    error_msg_8 = "The corresponding data files"
    with pytest.raises(RuntimeError, match=error_msg_8):
        all_data = ds.STL10Dataset(DATA_DIR, "all")
        all_data = all_data.map(operations=exception_func, input_columns=["image"], num_parallel_workers=1)
        for _ in all_data.__iter__():
            pass

    with pytest.raises(RuntimeError, match=error_msg_8):
        all_data = ds.STL10Dataset(DATA_DIR, "all")
        all_data = all_data.map(operations=vision.Decode(), input_columns=["image"], num_parallel_workers=1)
        for _ in all_data.__iter__():
            pass

    error_msg_9 = "does not exist or permission denied!"
    with pytest.raises(ValueError, match=error_msg_9):
        all_data = ds.STL10Dataset(WRONG_DIR, "all")
        for _ in all_data.__iter__():
            pass


def test_stl10_visualize(plot=False):
    """
    Feature: test_stl10_visualize.
    Description: visualize STL10Dataset results.
    Expectation: get correct number of data and plot them.
    """
    logger.info("Test STL10Dataset visualization")
    all_data = ds.STL10Dataset(DATA_DIR, "all", num_samples=2, shuffle=False)
    num_iter = 0
    image_list, label_list = [], []
    for item in all_data.create_dict_iterator(num_epochs=1, output_numpy=True):
        image = item["image"]
        label = item["label"]
        image_list.append(image)
        label_list.append("label {}".format(label))
        assert isinstance(image, np.ndarray)
        assert image.shape == (96, 96, 3)
        assert image.dtype == np.uint8
        assert label.dtype == np.int32
        num_iter += 1
    assert num_iter == 2
    if plot:
        visualize_dataset(image_list, label_list)


def test_stl10_usage():
    """
    Feature: test_stl10_usage.
    Description: validate STL10Dataset image readings.
    Expectation: get correct number of data.
    """
    logger.info("Test STL10Dataset usage flag")

    def test_config(usage, path=None):
        path = DATA_DIR if path is None else path
        try:
            data = ds.STL10Dataset(path, usage=usage, shuffle=False)
            num_rows = 0
            for _ in data.create_dict_iterator(num_epochs=1, output_numpy=True):
                num_rows += 1
        except (ValueError, TypeError, RuntimeError) as e:
            return str(e)
        return num_rows

    assert test_config("train") == 1
    assert test_config("test") == 1
    assert test_config("unlabeled") == 1
    assert test_config("train+unlabeled") == 2
    assert test_config("all") == 3

    assert "Input usage is not within the valid set of ['train', 'test', 'unlabeled', 'train+unlabeled', 'all']."\
             in test_config("invalid")
    assert "Argument usage with value ['list'] is not of type [<class 'str'>]" in test_config(["list"])

    # change this directory to the folder that contains all STL10 files.
    all_files_path = None
    # the following tests on the entire datasets.
    if all_files_path is not None:
        assert test_config("train", all_files_path) == 1

        assert ds.STL10Dataset(all_files_path, usage="train").get_dataset_size() == 1


if __name__ == '__main__':
    test_stl10_content_check()
    test_stl10_basic()
    test_stl10_sequential_sampler()
    test_stl10_exception()
    test_stl10_visualize(plot=True)
    test_stl10_usage()