How to use multiple_of method in Lemoncheesecake

Best Python code snippet using lemoncheesecake

test_crop_size_multiple.py

Source:test_crop_size_multiple.py

1#  ------------------------------------------------------------------------------------------2#  Copyright (c) Microsoft Corporation. All rights reserved.3#  Licensed under the MIT License (MIT). See LICENSE in the repo root for license information.4#  ------------------------------------------------------------------------------------------5from typing import Any, Optional6import pytest7import torch8from InnerEye.Common.type_annotations import IntOrTuple3, TupleInt39from InnerEye.ML.config import ModelArchitectureConfig, SegmentationModelBase10from InnerEye.ML.models.architectures.base_model import BaseSegmentationModel, CropSizeConstraints11from InnerEye.ML.models.architectures.unet_3d import UNet3D12from InnerEye.ML.utils.model_util import build_net13class SimpleModel(BaseSegmentationModel):14    def __init__(self, input_channels: int, channels: list, n_classes: int, kernel_size: int,15                 crop_size_constraints: CropSizeConstraints = None):16        super().__init__(input_channels=input_channels, name="SimpleModel", crop_size_constraints=crop_size_constraints)17        self.channels = channels18        self.n_classes = n_classes19        self.kernel_size = kernel_size20        self.model = torch.nn.Sequential(21            torch.nn.Conv3d(input_channels, channels[0], kernel_size=self.kernel_size),22            torch.nn.ConvTranspose3d(channels[0], n_classes, kernel_size=self.kernel_size)23        )24    def forward(self, x: Any) -> Any:  # type: ignore25        x = self.model(x)26        return x27def test_crop_size_constraints() -> None:28    """29    Test the basic logic to validate a crop size inside of a CropSizeConstraints instance.30    """31    def check(crop_size: TupleInt3,32              multiple_of: Optional[IntOrTuple3] = None,33              minimum: Optional[IntOrTuple3] = None) -> CropSizeConstraints:34        constraints = CropSizeConstraints(minimum_size=minimum, multiple_of=multiple_of)35        constraints.validate(crop_size)36        return constraints37    # crop_size_multiple == 1: Any crop size is allowed.38    c = check((9, 10, 11), multiple_of=1)39    # If a scalar multiple_of is used, it should be stored expanded along dimensions40    assert c.multiple_of == (1, 1, 1)41    # Using a tuple multiple_of: Crop is twice as large as multiple_of in each dimension, this is hence valid.42    c = check((10, 12, 14), multiple_of=(5, 6, 7))43    assert c.multiple_of == (5, 6, 7)44    # Minimum size has not been provided, should default to multiple_of.45    assert c.minimum_size == (5, 6, 7)46    # Try with a couple more common crop sizes47    check((32, 64, 64), multiple_of=16)48    check((1, 64, 64), multiple_of=(1, 16, 16))49    # Crop size is at the allowed minimum: This is valid50    check((3, 4, 5), multiple_of=(3, 4, 5))51    # Provide a scalar minimum: Should be stored expanded into 3 dimensions52    c = check((9, 10, 11), minimum=2)53    assert c.minimum_size == (2, 2, 2)54    assert c.multiple_of is None55    # Provide a tuple minimum56    c = check((9, 10, 11), minimum=(5, 6, 7))57    assert c.minimum_size == (5, 6, 7)58    # A crop size at exactly the minimum is valid59    check((5, 6, 7), minimum=(5, 6, 7))60    # Checking for minimum and multiple at the same time61    check((9, 10, 11), minimum=1, multiple_of=1)62    check((10, 12, 14), minimum=(5, 6, 7), multiple_of=2)63    def assert_fails(crop_size: TupleInt3,64                     multiple_of: Optional[IntOrTuple3] = None,65                     minimum: Optional[IntOrTuple3] = None) -> None:66        with pytest.raises(ValueError) as err:67            check(crop_size, multiple_of, minimum)68        assert str(crop_size) in str(err)69        assert "Crop size is not valid" in str(err)70    # Crop size is not a multiple of 2 in dimensions 0 and 171    assert_fails((3, 4, 5), 2)72    # Crop size is not a multiple of 6 in dimension 273    assert_fails((3, 4, 5), (3, 4, 6))74    assert_fails((16, 16, 200), 16)75    # Crop size is too small76    assert_fails((1, 2, 3), (10, 10, 10))77    assert_fails((10, 20, 30), multiple_of=10, minimum=20)78    # Minimum size must be at least as large as multiple_of:79    with pytest.raises(ValueError) as err:80        CropSizeConstraints(minimum_size=10, multiple_of=20, num_dimensions=2)81    assert "minimum size must be at least as large as the multiple_of" in str(err)82    assert "(10, 10)" in str(err)83    assert "(20, 20)" in str(err)84    # Check that num_dimensions is working as expected (though not presently used in the codebase)85    c = CropSizeConstraints(minimum_size=30, multiple_of=10, num_dimensions=2)86    assert c.multiple_of == (10, 10)87    assert c.minimum_size == (30, 30)88def test_crop_size_constructor() -> None:89    """90    Test error handling in the constructor of CropSizeConstraints91    """92    # Minimum size is given as 3-tuple, but working with 2 dimensions:93    with pytest.raises(ValueError) as err:94        CropSizeConstraints(minimum_size=(1, 2, 3), num_dimensions=2)95    assert "must have length 2" in str(err)96    # Minimum size and multiple_of are not compatible:97    with pytest.raises(ValueError) as err:98        CropSizeConstraints(minimum_size=(1, 2, 3), multiple_of=16)99    assert "The minimum size must be at least as large" in str(err)100def test_crop_size_check_in_model() -> None:101    """102    Test that crop size checks are correctly integrated and called from within BaseModel103    """104    def create_model(crop_size: TupleInt3, multiple_of: IntOrTuple3) -> BaseSegmentationModel:105        model = SimpleModel(1, [1], 2, 2, crop_size_constraints=CropSizeConstraints(multiple_of=multiple_of))106        model.validate_crop_size(crop_size)107        return model108    # crop_size_multiple == 1: Any crop size is allowed.109    m = create_model((9, 10, 11), 1)110    assert m.crop_size_constraints.multiple_of == (1, 1, 1)111    m = create_model((10, 12, 14), (5, 6, 7))112    assert m.crop_size_constraints.multiple_of == (5, 6, 7)113    def assert_fails(crop_size: TupleInt3, crop_size_multiple: IntOrTuple3) -> None:114        with pytest.raises(ValueError) as err:115            create_model(crop_size, crop_size_multiple)116        assert "Crop size is not valid" in str(err)117    # Crop size is not a multiple of 2 in dimensions 0 and 1118    assert_fails((3, 4, 5), 2)119    # Crop size constraints field should be generated as not empty if not provided in the120    # constructor - this simplifies code in the inference pipeline.121    model = SimpleModel(1, [1], 2, 2)122    assert model.crop_size_constraints is not None123    assert model.crop_size_constraints.minimum_size == (1, 1, 1)124    assert model.crop_size_constraints.multiple_of == (1, 1, 1)125    model.validate_crop_size((1, 1, 1))126def test_crop_size_multiple_in_build_net() -> None:127    """128    Tests if the the crop_size validation is really called in the model creation code129    """130    config = SegmentationModelBase(architecture=ModelArchitectureConfig.UNet3D,131                                   image_channels=["ct"],132                                   feature_channels=[1],133                                   kernel_size=3,134                                   crop_size=(17, 16, 16),135                                   should_validate=False)136    # Crop size of 17 in dimension 0 is invalid for a UNet3D, should be multiple of 16.137    # This should raise a ValueError.138    with pytest.raises(ValueError) as ex:139        build_net(config)140    assert "Training crop size: Crop size is not valid" in str(ex)141    config.crop_size = (16, 16, 16)142    config.test_crop_size = (17, 18, 19)143    with pytest.raises(ValueError) as ex:144        build_net(config)145    assert "Test crop size: Crop size is not valid" in str(ex)146def crop_size_multiple_unet3d(crop_size: TupleInt3,147                              downsampling_factor: IntOrTuple3,148                              num_downsampling_paths: int,149                              expected_crop_size_multiple: TupleInt3) -> None:150    model = UNet3D(name="",151                   input_image_channels=1,152                   initial_feature_channels=32,153                   num_classes=2,154                   kernel_size=3,155                   downsampling_factor=downsampling_factor,156                   num_downsampling_paths=num_downsampling_paths)157    assert model.crop_size_constraints.multiple_of == expected_crop_size_multiple158    model.validate_crop_size(crop_size)159def test_crop_size_multiple_unet3d() -> None:160    # Standard UNet with downsampling 2, 4 downsampling stages: Minimum crop is 16161    crop_size_multiple_unet3d(crop_size=(16, 64, 64), downsampling_factor=2, num_downsampling_paths=4,162                              expected_crop_size_multiple=(16, 16, 16))163    # Only 3 downsampling stages: each reduces by a factor 2, hence minimum crop size should be 8164    crop_size_multiple_unet3d(crop_size=(16, 64, 64), downsampling_factor=2, num_downsampling_paths=3,165                              expected_crop_size_multiple=(8, 8, 8))166    # UNet3D as used in UNet2D: No down-sampling in Z direction. Any crop size in Z is allowed, crop size167    # at least 16 in X and Y168    crop_size_multiple_unet3d(crop_size=(16, 64, 64), downsampling_factor=(1, 2, 2), num_downsampling_paths=4,169                              expected_crop_size_multiple=(1, 16, 16))170    # Non-standard downsampling factor of 3, 2 downsampling stages: Minimum crop is 3*3 = 9171    crop_size_multiple_unet3d(crop_size=(9, 9, 9), downsampling_factor=3, num_downsampling_paths=2,172                              expected_crop_size_multiple=(9, 9, 9))173def test_restrict_crop_size_too_small() -> None:174    """Test the modification of crop sizes when the image size is below the minimum."""175    shape = (10, 30, 40)176    crop_size = (20, 40, 20)177    stride = (10, 20, 20)178    constraint = CropSizeConstraints(multiple_of=16)179    with pytest.raises(ValueError) as e:180        constraint.restrict_crop_size_to_image(shape, crop_size, stride)181    # Error message should contain the actual image size182    assert str(shape) in e.value.args[0]183    assert "16" in e.value.args[0]184def check_restrict_crop(constraint: CropSizeConstraints,185                        shape: TupleInt3,186                        crop_size: TupleInt3,187                        stride: TupleInt3,188                        expected_crop: TupleInt3,189                        expected_stride: TupleInt3) -> None:190    (crop_new, stride_new) = constraint.restrict_crop_size_to_image(shape, crop_size, stride)191    assert crop_new == expected_crop192    assert stride_new == expected_stride193    # Stride and crop must be integer tuples194    assert isinstance(crop_new[0], int)195    assert isinstance(stride_new[0], int)196def test_restrict_crop_size() -> None:197    """Test the modification of crop sizes when the image size is smaller than the crop."""198    shape = (20, 35, 40)199    crop_size = (25, 40, 20)200    stride = (10, 20, 20)201    constraint = CropSizeConstraints(multiple_of=16)202    # Expected new crop size is the elementwise minimum of crop_size and shape,203    # rounded up to the nearest multiple of 16204    expected_crop = (32, 48, 32)205    # Stride should maintain (elementwise) the same ratio to crop as before206    expected_stride = (12, 24, 32)207    check_restrict_crop(constraint, shape, crop_size, stride, expected_crop, expected_stride)208def test_restrict_crop_size_tuple() -> None:209    """Test the modification of crop sizes when the image size is smaller than the crop,210    and the crop_multiple is a tuple with element-wise multiples."""211    shape = (20, 35, 40)212    crop_size = (25, 40, 20)213    stride = (10, 20, 20)214    constraint = CropSizeConstraints(multiple_of=(1, 16, 16))215    # Expected new crop size is the elementwise minimum of crop_size and shape,216    # rounded down to the nearest multiple of 16, apart from dimension 0217    expected_crop = (20, 48, 32)218    expected_stride = (8, 24, 32)219    check_restrict_crop(constraint, shape, crop_size, stride, expected_crop, expected_stride)220def test_restrict_crop_size_large_image() -> None:221    """Test the modification of crop sizes when the image size is larger than the crop:222    The crop and stride should be returned unchanged."""223    shape = (30, 50, 50)224    crop_size = (20, 40, 40)225    stride = crop_size226    constraint = CropSizeConstraints(multiple_of=1)227    expected_crop = crop_size228    expected_stride = stride229    check_restrict_crop(constraint, shape, crop_size, stride, expected_crop, expected_stride)230    constraint2 = CropSizeConstraints(multiple_of=1, minimum_size=999)231    with pytest.raises(ValueError) as err:232        check_restrict_crop(constraint2, shape, crop_size, stride, expected_crop, expected_stride)233    assert "at least a size of" in str(err)234    assert "999" in str(err)235def test_restrict_crop_size_uneven() -> None:236    """237    Test a case when the image is larger than the crop in Z, but not in X and Y.238    """239    shape = (20, 30, 30)240    crop_size = (10, 60, 60)241    stride = crop_size242    constraint = CropSizeConstraints(multiple_of=1)243    expected_crop = (10, 30, 30)244    expected_stride = (10, 30, 30)...

game_list.py

Source:game_list.py

1import collections2import functools3import gspread_asyncio4from oauth2client.service_account import ServiceAccountCredentials5import re6import common7## Range & co8@functools.total_ordering9class _Least:10    # Default __eq__ is fine assuming there is only one instance11    def __gt__(self, other):12        return False13least = _Least()14@functools.total_ordering15class _Greatest:16    # Default __eq__ is fine assuming there is only one instance17    def __lt__(self, other):18        return False19greatest = _Greatest()20def _round_low_to_multiple(low, multiple_of):21    return low if low is least else low + (-low % multiple_of)22def _round_high_to_multiple(high, multiple_of):23    return high if high is greatest else high - (high % multiple_of)24class Range:25    def __init__(self, low, high, multiple_of=1):26        assert (low is least) or isinstance(low, int)27        assert (high is greatest) or isinstance(high, int)28        assert isinstance(multiple_of, int)29        self.low = _round_low_to_multiple(low, multiple_of)30        self.high = _round_high_to_multiple(high, multiple_of)31        self.multiple_of = multiple_of32    def __bool__(self):33        return self.low <= self.high34    def __contains__(self, x):35        return (self.low <= x <= self.high) and ((x % self.multiple_of) == 0)36    def __str__(self):37        if not self:38            return 'none'39        if self.low is least:40            if self.high is greatest:41                s = 'any'42            else:43                s = f'up to {self.high}'44        elif self.high is greatest:45            s = f'{self.low}+'46        elif self.low == self.high:47            return f'{self.low}'48        else:49            s = f'{self.low}..{self.high}'50        if self.multiple_of == 1:51            return s52        if self.multiple_of == 2:53            return f'{s} even'54        return f'{s} multiple of {self.multiple_of}'55    def simplified(self, implicit_low, implicit_high):56        assert (implicit_low is least) or isinstance(implicit_low, int)57        assert (implicit_high is greatest) or isinstance(implicit_high, int)58        implicit_low = _round_low_to_multiple(implicit_low, self.multiple_of)59        implicit_high = _round_high_to_multiple(implicit_high, self.multiple_of)60        low = max(self.low, implicit_low)61        high = min(self.high, implicit_high)62        if low >= high:63            # Either empty or one number in range64            return Range(low, high)65        return Range(66            least if low == implicit_low else low,67            greatest if high == implicit_high else high,68            self.multiple_of)69## Parsing70_num_re = re.compile('[0-9]+')71def _nums(s):72    for m in _num_re.finditer(s):73        yield int(m.group())74def _max_num(s):75    return max(_nums(s), default=None)76_num_plus_re = re.compile(r'([0-9]+)\+')77def _num_range(s):78    multiple_of = 2 if 'even' in s.lower() else 179    if m := _num_plus_re.search(s):80        return Range(int(m.group(1)), greatest, multiple_of)81    if nums := list(_nums(s)):82        return Range(min(nums), max(nums), multiple_of)83    return None84_Field = collections.namedtuple('_Field',85    'name heading_re parse sub_heading_row required',86    defaults=(lambda s: s, None, False))87_fields = [88    _Field('title', re.compile('title', re.IGNORECASE), required=True),89    _Field('platform', re.compile('platform', re.IGNORECASE)),90    _Field('max_players', re.compile('max.+player', re.IGNORECASE), _max_num),91    _Field('good_players', re.compile('good.+player', re.IGNORECASE), _num_range),92    _Field('owns', re.compile('who.+owns', re.IGNORECASE), bool, sub_heading_row=2)]93_heading_row = 094_data_begin_row = 395Game = collections.namedtuple('Game', (field.name for field in _fields))96def _fixup(game):97    if game.good_players is None:98        return game99    if good_players := game.good_players.simplified(1,100            greatest if game.max_players is None else game.max_players):101        return game._replace(good_players=good_players)102    # Assume we didn't parse the good players field properly...103    return game._replace(good_players=None)104class _Loc:105    def __init__(self, field):106        self.field = field107        self.begin_col = None108        # self.end_col set once found109        # self.sub_headings set at end of _get_locs110def _get_locs(rows):111    locs = {field.name: _Loc(field) for field in _fields}112    cont_loc = None113    for col, heading in enumerate(rows[_heading_row]):114        if heading:115            cont_loc = None116            for loc in locs.values():117                heading_re = loc.field.heading_re118                if heading_re.search(heading):119                    if loc.begin_col is not None:120                        raise common.Error(121                            f'I found multiple headings matching "{heading_re.pattern}" in the GL spreadsheet.')122                    loc.begin_col = col123                    loc.end_col = col + 1124                    if loc.field.sub_heading_row is not None:125                        cont_loc = loc126                    break127        elif cont_loc is not None:128            if rows[cont_loc.field.sub_heading_row][col]:129                cont_loc.end_col = col + 1130            else:131                cont_loc = None132    for loc in locs.values():133        if loc.begin_col is None:134            raise common.Error(135                f'I couldn\'t find a heading matching "{loc.field.heading_re.pattern}" in the GL spreadsheet.')136        if loc.field.sub_heading_row is not None:137            loc.sub_headings = rows[loc.field.sub_heading_row][loc.begin_col:loc.end_col]138            seen = set()139            for sub_heading in loc.sub_headings:140                if sub_heading in seen:141                    raise common.Error(142                        'There are multiple columns in the GL spreadsheet under '143                        f'{rows[_heading_row][loc.begin_col]} with the same '144                        f'sub-heading ({sub_heading}).')145                seen.add(sub_heading)146    return locs147class _MissingRequiredField(Exception):148    pass149def _parse_field(loc, row):150    values = []151    for s in row[loc.begin_col:loc.end_col]:152        if loc.field.required and not s:153            raise _MissingRequiredField()154        values.append(loc.field.parse(s))155    return values[0] if loc.field.sub_heading_row is None else dict(zip(loc.sub_headings, values))156GameList = collections.namedtuple('GameList', 'names games')157def _parse(rows):158    rows = [[s.strip() for s in row] for row in rows]159    locs = _get_locs(rows)160    names = set(locs['owns'].sub_headings)161    games = []162    for row in rows[_data_begin_row:]:163        try:164            values = {loc.field.name: _parse_field(loc, row) for loc in locs.values()}165        except _MissingRequiredField:166            continue167        games.append(_fixup(Game(**values)))168    return GameList(names, games)169## Google Sheets stuff170def _make_gsheets_client_manager(gsheets_creds):171    def get_creds():172        return ServiceAccountCredentials.from_json_keyfile_name(173            gsheets_creds, ['https://www.googleapis.com/auth/spreadsheets.readonly'])174    return gspread_asyncio.AsyncioGspreadClientManager(get_creds)175async def _fetch_gsheets_rows(cm, id):176    c = await cm.authorize()177    ss = await c.open_by_key(id)178    ws = await ss.get_worksheet(0)179    return await ws.get_all_values()180## Top level181class Session:182    def __init__(self, gsheets_creds):183        self.__cm = _make_gsheets_client_manager(gsheets_creds)184    async def fetch(self, id):185        rows = await _fetch_gsheets_rows(self.__cm, id)...

Ostroumov_DZ_6-1.py

Source:Ostroumov_DZ_6-1.py

1"""21. ÐÐ¾Ð´ÑÑÐ¸ÑÐ°ÑÑ, ÑÐºÐ¾Ð»ÑÐºÐ¾ Ð±ÑÐ»Ð¾ Ð²ÑÐ´ÐµÐ»ÐµÐ½Ð¾ Ð¿Ð°Ð¼ÑÑÐ¸ Ð¿Ð¾Ð´ Ð¿ÐµÑÐµÐ¼ÐµÐ½Ð½ÑÐµ Ð² ÑÐ°Ð½ÐµÐµ ÑÐ°Ð·ÑÐ°Ð±Ð¾ÑÐ°Ð½Ð½ÑÑ3Ð¿ÑÐ¾Ð³ÑÐ°Ð¼Ð¼Ð°Ñ Ð² ÑÐ°Ð¼ÐºÐ°Ñ Ð¿ÐµÑÐ²ÑÑ ÑÑÐµÑ ÑÑÐ¾ÐºÐ¾Ð².4ÐÑÐ¾Ð°Ð½Ð°Ð»Ð¸Ð·Ð¸ÑÐ¾Ð²Ð°ÑÑ ÑÐµÐ·ÑÐ»ÑÑÐ°Ñ Ð¸ Ð¾Ð¿ÑÐµÐ´ÐµÐ»Ð¸ÑÑ Ð¿ÑÐ¾Ð³ÑÐ°Ð¼Ð¼Ñ Ñ Ð½Ð°Ð¸Ð±Ð¾Ð»ÐµÐµ ÑÑÑÐµÐºÑÐ¸Ð²Ð½ÑÐ¼ Ð¸ÑÐ¿Ð¾Ð»ÑÐ·Ð¾Ð²Ð°Ð½Ð¸ÐµÐ¼ Ð¿Ð°Ð¼ÑÑÐ¸.5ÐÑÐ¸Ð¼ÐµÑÐ°Ð½Ð¸Ðµ: ÐÐ»Ñ Ð°Ð½Ð°Ð»Ð¸Ð·Ð° Ð²Ð¾Ð·ÑÐ¼Ð¸ÑÐµ Ð»ÑÐ±ÑÐµ 1-3 Ð²Ð°ÑÐ¸Ñ Ð¿ÑÐ¾Ð³ÑÐ°Ð¼Ð¼Ñ Ð¸Ð»Ð¸ Ð½ÐµÑÐºÐ¾Ð»ÑÐºÐ¾ Ð²Ð°ÑÐ¸Ð°Ð½ÑÐ¾Ð² ÐºÐ¾Ð´Ð° Ð´Ð»Ñ Ð¾Ð´Ð½Ð¾Ð¹ Ð¸ ÑÐ¾Ð¹ Ð¶Ðµ Ð·Ð°Ð´Ð°ÑÐ¸.6Ð ÐµÐ·ÑÐ»ÑÑÐ°ÑÑ Ð°Ð½Ð°Ð»Ð¸Ð·Ð° Ð²ÑÑÐ°Ð²ÑÑÐµ Ð² Ð²Ð¸Ð´Ðµ ÐºÐ¾Ð¼Ð¼ÐµÐ½ÑÐ°ÑÐ¸ÐµÐ² Ðº ÐºÐ¾Ð´Ñ.7Ð¢Ð°ÐºÐ¶Ðµ ÑÐºÐ°Ð¶Ð¸ÑÐµ Ð² ÐºÐ¾Ð¼Ð¼ÐµÐ½ÑÐ°ÑÐ¸ÑÑ Ð²ÐµÑÑÐ¸Ñ Python Ð¸ ÑÐ°Ð·ÑÑÐ´Ð½Ð¾ÑÑÑ Ð²Ð°ÑÐµÐ¹ ÐÐ¡.8/// Ð²ÐµÑÑÐ¸Ñ Python 3.99/// Windows 10 pro x6410/// 3-8. ÐÐ°ÑÑÐ¸ÑÐ° 5x4 Ð·Ð°Ð¿Ð¾Ð»Ð½ÑÐµÑÑÑ Ð²Ð²Ð¾Ð´Ð¾Ð¼ Ñ ÐºÐ»Ð°Ð²Ð¸Ð°ÑÑÑÑ ÐºÑÐ¾Ð¼Ðµ Ð¿Ð¾ÑÐ»ÐµÐ´Ð½Ð¸Ñ ÑÐ»ÐµÐ¼ÐµÐ½ÑÐ¾Ð² ÑÑÑÐ¾Ðº.11ÐÑÐ¾Ð³ÑÐ°Ð¼Ð¼Ð° Ð´Ð¾Ð»Ð¶Ð½Ð° Ð²ÑÑÐ¸ÑÐ»ÑÑÑ ÑÑÐ¼Ð¼Ñ Ð²Ð²ÐµÐ´ÐµÐ½Ð½ÑÑ ÑÐ»ÐµÐ¼ÐµÐ½ÑÐ¾Ð² ÐºÐ°Ð¶Ð´Ð¾Ð¹ ÑÑÑÐ¾ÐºÐ¸ Ð¸ Ð·Ð°Ð¿Ð¸ÑÑÐ²Ð°ÑÑ ÐµÐµ Ð² Ð¿Ð¾ÑÐ»ÐµÐ´Ð½ÑÑ ÑÑÐµÐ¹ÐºÑ ÑÑÑÐ¾ÐºÐ¸.12Ð ÐºÐ¾Ð½ÑÐµ ÑÐ»ÐµÐ´ÑÐµÑ Ð²ÑÐ²ÐµÑÑÐ¸ Ð¿Ð¾Ð»ÑÑÐµÐ½Ð½ÑÑ Ð¼Ð°ÑÑÐ¸ÑÑ.13"""14import random15import sys16def m_string(a):17    """18    ÑÐ¾ÑÐ¼Ð¸ÑÑÐµÑ Ð¼Ð°ÑÑÐ¸Ð² Ð½Ð° Ð·Ð°Ð´Ð°Ð½Ð½Ð¾Ðµ ÐºÐ¾Ð»Ð¸ÑÐµÑÑÐ²Ð¾ ÑÐ»ÐµÐ¼ÐµÐ½ÑÐ¾Ð²19    :param a: Ð´Ð»Ð¸Ð½Ð½Ð° ÑÑÐµÐ±ÑÐµÐ¼Ð¾Ð³Ð¾ Ð¼Ð°ÑÑÐ¸Ð²Ð°20    :return: Ð²Ð¾Ð·ÑÐ°ÑÐ°ÐµÑ Ð¼Ð°ÑÑÐ¸Ð²21    """22    array = [0] * a23    for i in range(a):24        array[i] = int(random.randint(1, 100))25    return array26def counter(a):27    """28    ÑÑÐ¸ÑÐ°ÐµÑ ÑÑÐ¼Ð¼Ñ ÑÐ»ÐµÐ¼ÐµÐ½ÑÐ¾Ð² Ð¼Ð°ÑÑÐ¸Ð²Ð° Ð¸ Ð·Ð°Ð¿Ð¸ÑÑÐ²Ð°ÐµÑ ÑÐµÐ·ÑÐ»ÑÑÐ°Ñ Ð² ÐºÐ¾Ð½ÐµÑ Ð¼Ð°ÑÑÐ¸Ð²Ð°29    :param a: Ð¿ÐµÑÐµÐ´Ð°ÐµÑÑÑ Ð¼Ð°ÑÑÐ¸Ð²30    :return: Ð²Ð¾Ð·ÑÐ°ÑÐ°ÐµÑÑÑ Ð¼Ð°ÑÑÐ¸Ð² Ñ Ð¿Ð¾ÑÑÐ¸ÑÐ°Ð½Ð½Ð¾Ð¹ ÑÑÐ¼Ð¼Ð¾Ð¹ ÑÐ»ÐµÐ¼ÐµÐ½ÑÐ¾Ð² Ð¼Ð°ÑÑÐ¸Ð²Ð°31    """32    summ = 033    for item in a:34        summ += int(item)35    return a.append(summ)36m_string_1 = m_string(3)37counter(m_string_1)38result = [m_string_1]39result_3 = sys.getsizeof(result) + sys.getsizeof(m_string_1)40print(f'{sys.getsizeof(result) = }, {sys.getsizeof(m_string_1) = },\n'41      f'Ð¾Ð±ÑÐµÐµ ÐºÐ¾Ð»Ð¸ÑÐµÑÑÐ²Ð¾ Ð·Ð°ÑÑÐ°ÑÐµÐ½Ð½Ð¾Ð¹ Ð¿Ð°Ð¼ÑÑÐ¸: {result_3}\n \n \n')42"""43sys.getsizeof(result) = 64, sys.getsizeof(m_string_1) = 120,44Ð¾Ð±ÑÐµÐµ ÐºÐ¾Ð»Ð¸ÑÐµÑÑÐ²Ð¾ Ð·Ð°ÑÑÐ°ÑÐµÐ½Ð½Ð¾Ð¹ Ð¿Ð°Ð¼ÑÑÐ¸: 18445/// 3 - 1. Ð Ð´Ð¸Ð°Ð¿Ð°Ð·Ð¾Ð½Ðµ Ð½Ð°ÑÑÑÐ°Ð»ÑÐ½ÑÑ ÑÐ¸ÑÐµÐ» Ð¾Ñ 2 Ð´Ð¾ 99 Ð¾Ð¿ÑÐµÐ´ÐµÐ»Ð¸ÑÑ,46ÑÐºÐ¾Ð»ÑÐºÐ¾ Ð¸Ð· Ð½Ð¸Ñ ÐºÑÐ°ÑÐ½Ñ ÐºÐ°Ð¶Ð´Ð¾Ð¼Ñ Ð¸Ð· ÑÐ¸ÑÐµÐ» Ð² Ð´Ð¸Ð°Ð¿Ð°Ð·Ð¾Ð½Ðµ Ð¾Ñ 2 Ð´Ð¾ 9.47"""48nums = [n for n in range(2, 100)]49multiple_of = {"Ð´Ð²Ð°": 0, "ÑÑÐ¸": 0, "ÑÐµÑÑÑÐµ": 0, "Ð¿ÑÑÑ": 0, "ÑÐµÑÑÑ": 0, "ÑÐµÐ¼Ñ": 0, "Ð²Ð¾ÑÐµÐ¼Ñ": 0, "Ð´ÐµÐ²ÑÑÑ": 0}50# Ð´Ð»Ñ Ð°Ð½Ð°Ð»Ð¸Ð·Ð° Ð²Ð²Ð¾Ð¶Ñ Ð´Ð¾Ð¿Ð¾Ð»Ð½Ð¸ÐµÑÐµÐ»ÑÐ½ÑÐµ Ð¿ÐµÑÐµÐ¼ÐµÐ½Ð½ÑÐµ:51two = 052three = 053four = 054five = 055six = 056seven = 057eight = 058nine = 059for item in nums:60    if item % 2 == 0:61        multiple_of["Ð´Ð²Ð°"] += 162        two += 163    if item % 3 == 0:64        multiple_of["ÑÑÐ¸"] += 165        three += 166    if item % 4 == 0:67        multiple_of["ÑÐµÑÑÑÐµ"] += 168        four += 169    if item % 5 == 0:70        multiple_of["Ð¿ÑÑÑ"] += 171        five += 172    if item % 6 == 0:73        multiple_of["ÑÐµÑÑÑ"] += 174        six += 175    if item % 7 == 0:76        multiple_of["ÑÐµÐ¼Ñ"] += 177        seven += 178    if item % 8 == 0:79        multiple_of["Ð²Ð¾ÑÐµÐ¼Ñ"] += 180        eight += 181    if item % 9 == 0:82        multiple_of["Ð´ÐµÐ²ÑÑÑ"] += 183        nine += 184result_1 = sys.getsizeof(nums) + sys.getsizeof(multiple_of)85result_2 = two + three + four + five + six + seven + eight + nine + sys.getsizeof(nums)86print(f'{sys.getsizeof(nums) = }, {sys.getsizeof(multiple_of) = }, Ð²ÑÐµÐ³Ð¾ {result_1}')87print(f'ÐºÐ°Ð¶Ð´Ð°Ñ "Ð¸Ð½Ð´Ð¸Ð²Ð´ÑÐ°Ð»ÑÐ½Ð°Ñ" Ð¿ÐµÑÐµÐ¼ÐµÐ½Ð½Ð°Ñ Ð±ÑÐ´ÐµÑ Ð·Ð°Ð½Ð¸Ð¼Ð°ÑÑ  {two} Ð¿Ð°Ð¼ÑÑÐ¸')88print(f'Ð±ÐµÐ· ÑÐ»Ð¾Ð²Ð°ÑÑ, Ñ Ð¸Ð½Ð´Ð¸Ð²Ð¸Ð´ÑÐ°Ð»ÑÐ½ÑÐ¼Ð¸ Ð¿ÐµÑÐµÐ¼ÐµÐ½Ð½ÑÐ¼Ð¸: {result_2}')89'''90sys.getsizeof(nums) = 920, sys.getsizeof(multiple_of) = 360, Ð²ÑÐµÐ³Ð¾ 128091ÐºÐ°Ð¶Ð´Ð°Ñ "Ð¸Ð½Ð´Ð¸Ð²Ð´ÑÐ°Ð»ÑÐ½Ð°Ñ" Ð¿ÐµÑÐµÐ¼ÐµÐ½Ð½Ð°Ñ Ð±ÑÐ´ÐµÑ Ð·Ð°Ð½Ð¸Ð¼Ð°ÑÑ  49 Ð¿Ð°Ð¼ÑÑÐ¸92Ð±ÐµÐ· ÑÐ»Ð¾Ð²Ð°ÑÑ, Ñ Ð¸Ð½Ð´Ð¸Ð²Ð¸Ð´ÑÐ°Ð»ÑÐ½ÑÐ¼Ð¸ Ð¿ÐµÑÐµÐ¼ÐµÐ½Ð½ÑÐ¼Ð¸: 1098...

transforms_resize_modulo_pad_crop.py

Source:transforms_resize_modulo_pad_crop.py

1import collections2import functools3from numbers import Number4from typing import Union, Optional5import numpy as np6from .transforms import CriteriaFn7from ..transforms import transforms8from ..transforms import crop9from ..basic_typing import ShapeX10from ..utils import batch_pad_minmax_joint11from typing_extensions import Literal12def _transform_resize_modulo_crop_pad(13        features_names,14        batch,15        multiple_of,16        mode,17        padding_mode,18        padding_constant_value):19    resized_arrays = []20    arrays = [batch[name] for name in features_names]21    for a in arrays[1:]:22        assert len(a.shape) == len(arrays[0].shape)23        assert a.shape[2:] == arrays[0].shape[2:], f'selected tensors MUST have the same size. ' \24                                                   f'Expected={a.shape}, got={arrays[0].shape}'25    if len(features_names) > 0:26        shape = arrays[0].shape[2:]27        if isinstance(multiple_of, Number):28            multiple_of = [multiple_of] * (len(shape))29        multiple_of = np.asarray(multiple_of)30        assert len(shape) == len(multiple_of), 'expected multiple_of expressed as [D]HW shape!'31        shape_extra = arrays[0].shape[2:] % multiple_of32        if mode == 'pad':33            # padding is centered34            padding = multiple_of - shape_extra35            padding_left = padding // 236            padding_right = padding - padding_left37            padding_left = [0] + list(padding_left)38            padding_right = [0] + list(padding_right)39            resized_arrays = batch_pad_minmax_joint(40                arrays,41                padding_left,42                padding_right,43                padding_mode,44                padding_constant_value45            )46        elif mode == 'crop':47            s = list(shape - shape_extra)48            assert min(s) > 0, f'shape is too small! Use a different mode. ' \49                               f'Got={arrays[0].shape[2:]}, multiple_of={multiple_of}'50            resized_arrays = crop.transform_batch_random_crop_joint(arrays, [None] + s)51        else:52            raise NotImplementedError(f'mode={mode} is not handled!')53    new_batch = collections.OrderedDict(zip(features_names, resized_arrays))54    for feature_name, feature_value in batch.items():55        if feature_name not in features_names:56            # not in the transformed features, so copy the original value57            new_batch[feature_name] = feature_value58    return new_batch59class TransformResizeModuloCropPad(transforms.TransformBatchWithCriteria):60    """61    Resize tensors by padding or cropping the tensor so its shape is a multiple of a ``multiple_of``.62    This can be particularly helpful in encoder-decoder architecture with skip connection which can impose63    constraints on the input shape (e.g., the input must be a multiple of 32 pixels).64    Args:65        multiple_of: a sequence of size `len(array.shape)-2` such that shape % multiple_of == 0. To achieve this,66            the tensors will be padded or cropped.67        criteria_fn: function applied on each feature. If satisfied, the feature will be transformed, if not68            the original feature is returned69        padding_mode: `numpy.pad` mode. Currently supported are ('constant', 'edge', 'symmetric')70        mode: one of `crop`, `pad`. If `pad`, the selected tensors will be padded to achieve the71            size tensor.shape % multiple_of == 0. If `crop`, the selected tensors will be cropped72            instead with a randomly selected cropping position.73    Returns:74        dictionary with the selected tensors cropped or padded to the appropriate size75    """76    def __init__(77            self,78            multiple_of: Union[int, ShapeX],79            criteria_fn: Optional[CriteriaFn] = None,80            mode: Literal['crop', 'pad'] = 'crop',81            padding_mode: Literal['edge', 'constant', 'symmetric'] = 'constant',82            padding_constant_value: int = 0):83        if criteria_fn is None:84            criteria_fn = transforms.criteria_is_array_4_or_above85        super().__init__(86            criteria_fn=criteria_fn,87            transform_fn=functools.partial(88                _transform_resize_modulo_crop_pad,89                padding_mode=padding_mode,90                mode=mode,91                padding_constant_value=padding_constant_value,92                multiple_of=multiple_of)93         )...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.