mindspore2022/tests/ut/python/dataset/test_datasets_usps.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 USPS dataset operators
"""
import os
from typing import cast

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/testUSPSDataset"
WRONG_DIR = "../data/dataset/testMnistData"


def load_usps(path, usage):
    """
    load USPS data
    """
    assert usage in ["train", "test"]
    if usage == "train":
        data_path = os.path.realpath(os.path.join(path, "usps"))
    elif usage == "test":
        data_path = os.path.realpath(os.path.join(path, "usps.t"))

    with open(data_path, 'r') as f:
        raw_data = [line.split() for line in f.readlines()]
        tmp_list = [[x.split(':')[-1] for x in data[1:]] for data in raw_data]
        images = np.asarray(tmp_list, dtype=np.float32).reshape((-1, 16, 16, 1))
        images = ((cast(np.ndarray, images) + 1) / 2 * 255).astype(dtype=np.uint8)
        labels = [int(d[0]) - 1 for d in raw_data]
    return images, labels


def visualize_dataset(images, labels):
    """
    Helper function to visualize the dataset samples
    """
    num_samples = len(images)
    for i in range(num_samples):
        plt.subplot(1, num_samples, i + 1)
        plt.imshow(images[i].squeeze(), cmap=plt.cm.gray)
        plt.title(labels[i])
    plt.show()


def test_usps_content_check():
    """
    Validate USPSDataset image readings
    """
    logger.info("Test USPSDataset Op with content check")
    train_data = ds.USPSDataset(DATA_DIR, "train", num_samples=10, shuffle=False)
    images, labels = load_usps(DATA_DIR, "train")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label"
    for i, data in enumerate(train_data.create_dict_iterator(num_epochs=1, output_numpy=True)):
        for m in range(16):
            for n in range(16):
                assert (data["image"][m, n, 0] != 0 or images[i][m, n, 0] != 255) and \
                        (data["image"][m, n, 0] != 255 or images[i][m, n, 0] != 0)
                assert (data["image"][m, n, 0] == images[i][m, n, 0]) or\
                        (data["image"][m, n, 0] == images[i][m, n, 0] + 1) or\
                        (data["image"][m, n, 0] + 1 == images[i][m, n, 0])
        np.testing.assert_array_equal(data["label"], labels[i])
        num_iter += 1
    assert num_iter == 3

    test_data = ds.USPSDataset(DATA_DIR, "test", num_samples=3, shuffle=False)
    images, labels = load_usps(DATA_DIR, "test")
    num_iter = 0
    # in this example, each dictionary has keys "image" and "label"
    for i, data in enumerate(test_data.create_dict_iterator(num_epochs=1, output_numpy=True)):
        for m in range(16):
            for n in range(16):
                if (data["image"][m, n, 0] == 0 and images[i][m, n, 0] == 255) or\
                        (data["image"][m, n, 0] == 255 and images[i][m, n, 0] == 0):
                    assert False
                if (data["image"][m, n, 0] != images[i][m, n, 0]) and\
                        (data["image"][m, n, 0] != images[i][m, n, 0] + 1) and\
                        (data["image"][m, n, 0] + 1 != images[i][m, n, 0]):
                    assert False
        np.testing.assert_array_equal(data["label"], labels[i])
        num_iter += 1
    assert num_iter == 3


def test_usps_basic():
    """
    Validate USPSDataset
    """
    logger.info("Test USPSDataset Op")

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

    test_data = ds.USPSDataset(DATA_DIR, "test")
    num_iter = 0
    for _ in test_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 3

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

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

    # case 4: test batch with drop_remainder=False
    train_data = ds.USPSDataset(DATA_DIR, "train", num_samples=3)
    assert train_data.get_dataset_size() == 3
    assert train_data.get_batch_size() == 1
    train_data = train_data.batch(batch_size=2)  # drop_remainder is default to be False
    assert train_data.get_batch_size() == 2
    assert train_data.get_dataset_size() == 2

    num_iter = 0
    for _ in train_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 2

    # case 5: test batch with drop_remainder=True
    train_data = ds.USPSDataset(DATA_DIR, "train", num_samples=3)
    assert train_data.get_dataset_size() == 3
    assert train_data.get_batch_size() == 1
    train_data = train_data.batch(batch_size=2, drop_remainder=True)  # the rest of incomplete batch will be dropped
    assert train_data.get_dataset_size() == 1
    assert train_data.get_batch_size() == 2
    num_iter = 0
    for _ in train_data.create_dict_iterator(num_epochs=1):
        num_iter += 1
    assert num_iter == 1


def test_usps_exception():
    """
    Test error cases for USPSDataset
    """
    error_msg_3 = "num_shards is specified and currently requires shard_id as well"
    with pytest.raises(RuntimeError, match=error_msg_3):
        ds.USPSDataset(DATA_DIR, "train", num_shards=10)
        ds.USPSDataset(DATA_DIR, "test", num_shards=10)

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

    error_msg_5 = "Input shard_id is not within the required interval"
    with pytest.raises(ValueError, match=error_msg_5):
        ds.USPSDataset(DATA_DIR, "train", num_shards=5, shard_id=-1)
        ds.USPSDataset(DATA_DIR, "test", num_shards=5, shard_id=-1)
    with pytest.raises(ValueError, match=error_msg_5):
        ds.USPSDataset(DATA_DIR, "train", num_shards=5, shard_id=5)
        ds.USPSDataset(DATA_DIR, "test", num_shards=5, shard_id=5)
    with pytest.raises(ValueError, match=error_msg_5):
        ds.USPSDataset(DATA_DIR, "train", num_shards=2, shard_id=5)
        ds.USPSDataset(DATA_DIR, "test", num_shards=2, shard_id=5)

    error_msg_6 = "num_parallel_workers exceeds"
    with pytest.raises(ValueError, match=error_msg_6):
        ds.USPSDataset(DATA_DIR, "train", shuffle=False, num_parallel_workers=0)
        ds.USPSDataset(DATA_DIR, "test", shuffle=False, num_parallel_workers=0)
    with pytest.raises(ValueError, match=error_msg_6):
        ds.USPSDataset(DATA_DIR, "train", shuffle=False, num_parallel_workers=256)
        ds.USPSDataset(DATA_DIR, "test", shuffle=False, num_parallel_workers=256)
    with pytest.raises(ValueError, match=error_msg_6):
        ds.USPSDataset(DATA_DIR, "train", shuffle=False, num_parallel_workers=-2)
        ds.USPSDataset(DATA_DIR, "test", shuffle=False, num_parallel_workers=-2)

    error_msg_7 = "Argument shard_id"
    with pytest.raises(TypeError, match=error_msg_7):
        ds.USPSDataset(DATA_DIR, "train", num_shards=2, shard_id="0")
        ds.USPSDataset(DATA_DIR, "test", num_shards=2, shard_id="0")

    error_msg_8 = "invalid input shape"
    with pytest.raises(RuntimeError, match=error_msg_8):
        train_data = ds.USPSDataset(DATA_DIR, "train")
        train_data = train_data.map(operations=vision.Decode(), input_columns=["image"], num_parallel_workers=1)
        for _ in train_data.__iter__():
            pass

        test_data = ds.USPSDataset(DATA_DIR, "test")
        test_data = test_data.map(operations=vision.Decode(), input_columns=["image"], num_parallel_workers=1)
        for _ in test_data.__iter__():
            pass

    error_msg_9 = "failed to find USPS train data file"
    with pytest.raises(RuntimeError, match=error_msg_9):
        train_data = ds.USPSDataset(WRONG_DIR, "train")
        for _ in train_data.__iter__():
            pass
    error_msg_10 = "failed to find USPS test data file"
    with pytest.raises(RuntimeError, match=error_msg_10):
        test_data = ds.USPSDataset(WRONG_DIR, "test")
        for _ in test_data.__iter__():
            pass


def test_usps_visualize(plot=False):
    """
    Visualize USPSDataset results
    """
    logger.info("Test USPSDataset visualization")

    train_data = ds.USPSDataset(DATA_DIR, "train", num_samples=3, shuffle=False)
    num_iter = 0
    image_list, label_list = [], []
    for item in train_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 == (16, 16, 1)
        assert image.dtype == np.uint8
        assert label.dtype == np.uint32
        num_iter += 1
    assert num_iter == 3
    if plot:
        visualize_dataset(image_list, label_list)

    test_data = ds.USPSDataset(DATA_DIR, "test", num_samples=3, shuffle=False)
    num_iter = 0
    image_list, label_list = [], []
    for item in test_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 == (16, 16, 1)
        assert image.dtype == np.uint8
        assert label.dtype == np.uint32
        num_iter += 1
    assert num_iter == 3
    if plot:
        visualize_dataset(image_list, label_list)


def test_usps_usage():
    """
    Validate USPSDataset image readings
    """
    logger.info("Test USPSDataset usage flag")

    def test_config(usage, path=None):
        path = DATA_DIR if path is None else path
        try:
            data = ds.USPSDataset(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") == 3
    assert test_config("test") == 3

    assert "usage is not within the valid set of ['train', 'test', '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 USPS 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) == 3
        assert test_config("test", all_files_path) == 3
        assert ds.USPSDataset(all_files_path, usage="train").get_dataset_size() == 3
        assert ds.USPSDataset(all_files_path, usage="test").get_dataset_size() == 3


if __name__ == '__main__':
    test_usps_content_check()
    test_usps_basic()
    test_usps_exception()
    test_usps_visualize(plot=True)
    test_usps_usage()