move random apply order and choice ops

mindspore/dataset/transforms/py_transforms.py

@@ -17,7 +17,7 @@
 This module py_transforms is implemented basing on Python. It provides common
 operations including OneHotOp.
 """
-from .validators import check_one_hot_op, check_compose_list
+from .validators import check_one_hot_op, check_compose_list, check_random_apply, check_transforms_list
 from . import py_transforms_util as util
 
 
@@ -100,3 +100,104 @@ class Compose:
             lambda function, Lambda function that takes in an img to apply transformations on.
         """
         return util.compose(img, self.transforms)
+
+
+class RandomApply:
+    """
+    Randomly perform a series of transforms with a given probability.
+
+    Args:
+        transforms (list): List of transformations to apply.
+        prob (float, optional): The probability to apply the transformation list (default=0.5).
+
+    Examples:
+        >>> import mindspore.dataset.vision.py_transforms as py_vision
+        >>> from mindspore.dataset.transforms.py_transforms import Compose
+        >>>
+        >>> Compose([py_vision.Decode(),
+        >>>          py_vision.RandomApply(transforms_list, prob=0.6),
+        >>>          py_vision.ToTensor()])
+    """
+
+    @check_random_apply
+    def __init__(self, transforms, prob=0.5):
+        self.prob = prob
+        self.transforms = transforms
+
+    def __call__(self, img):
+        """
+        Call method.
+
+        Args:
+            img (PIL image): Image to be randomly applied a list transformations.
+
+        Returns:
+            img (PIL image), Transformed image.
+        """
+        return util.random_apply(img, self.transforms, self.prob)
+
+
+class RandomChoice:
+    """
+    Randomly select one transform from a series of transforms and applies that on the image.
+
+    Args:
+         transforms (list): List of transformations to be chosen from to apply.
+
+    Examples:
+        >>> import mindspore.dataset.vision.py_transforms as py_vision
+        >>> from mindspore.dataset.transforms.py_transforms import Compose, RandomChoice
+        >>>
+        >>> Compose([py_vision.Decode(),
+        >>>          RandomChoice(transforms_list),
+        >>>          py_vision.ToTensor()])
+    """
+
+    @check_transforms_list
+    def __init__(self, transforms):
+        self.transforms = transforms
+
+    def __call__(self, img):
+        """
+        Call method.
+
+        Args:
+            img (PIL image): Image to be applied transformation.
+
+        Returns:
+            img (PIL image), Transformed image.
+        """
+        return util.random_choice(img, self.transforms)
+
+
+class RandomOrder:
+    """
+    Perform a series of transforms to the input PIL image in a random order.
+
+    Args:
+        transforms (list): List of the transformations to apply.
+
+    Examples:
+        >>> import mindspore.dataset.vision.py_transforms as py_vision
+        >>> from mindspore.dataset.transforms.py_transforms import Compose
+        >>>
+        >>> Compose([py_vision.Decode(),
+        >>>          py_vision.RandomOrder(transforms_list),
+        >>>          py_vision.ToTensor()])
+    """
+
+    @check_transforms_list
+    def __init__(self, transforms):
+        self.transforms = transforms
+
+    def __call__(self, img):
+        """
+        Call method.
+
+        Args:
+            img (PIL image): Image to apply transformations in a random order.
+
+        Returns:
+            img (PIL image), Transformed image.
+        """
+        return util.random_order(img, self.transforms)

mindspore/dataset/transforms/py_transforms_util.py

@@ -15,7 +15,9 @@
 """
 Built-in py_transforms_utils functions.
 """
+import random
 import numpy as np
+
 from ..core.py_util_helpers import is_numpy
 
 
@@ -63,3 +65,53 @@ def one_hot_encoding(label, num_classes, epsilon):
         one_hot_label[index, label[index]] = 1
 
     return (1 - epsilon) * one_hot_label + epsilon / num_classes
+
+
+def random_order(img, transforms):
+    """
+    Applies a list of transforms in a random order.
+
+    Args:
+        img: Image to be applied transformations in a random order.
+        transforms (list): List of the transformations to be applied.
+
+    Returns:
+        img, Transformed image.
+    """
+    random.shuffle(transforms)
+    for transform in transforms:
+        img = transform(img)
+    return img
+
+
+def random_apply(img, transforms, prob):
+    """
+    Apply a list of transformation, randomly with a given probability.
+
+    Args:
+        img: Image to be randomly applied a list transformations.
+        transforms (list): List of transformations to be applied.
+        prob (float): The probability to apply the transformation list.
+
+    Returns:
+        img, Transformed image.
+    """
+    if prob < random.random():
+        return img
+    for transform in transforms:
+        img = transform(img)
+    return img
+
+
+def random_choice(img, transforms):
+    """
+    Random selects one transform from a list of transforms and applies that on the image.
+
+    Args:
+        img: Image to be applied transformation.
+        transforms (list): List of transformations to be chosen from to apply.
+
+    Returns:
+        img, Transformed image.
+    """
+    return random.choice(transforms)(img)

mindspore/dataset/transforms/validators.py

@@ -216,3 +216,34 @@ def check_compose_list(method):
         return method(self, *args, **kwargs)
 
     return new_method
+
+
+def check_random_apply(method):
+    """Wrapper method to check the parameters of random apply."""
+
+    @wraps(method)
+    def new_method(self, *args, **kwargs):
+        [transforms, prob], _ = parse_user_args(method, *args, **kwargs)
+        type_check(transforms, (list,), "transforms")
+
+        if prob is not None:
+            type_check(prob, (float, int,), "prob")
+            check_value(prob, [0., 1.], "prob")
+
+        return method(self, *args, **kwargs)
+
+    return new_method
+
+
+def check_transforms_list(method):
+    """Wrapper method to check the parameters of transform list."""
+
+    @wraps(method)
+    def new_method(self, *args, **kwargs):
+        [transforms], _ = parse_user_args(method, *args, **kwargs)
+
+        type_check(transforms, (list,), "transforms")
+
+        return method(self, *args, **kwargs)
+
+    return new_method

mindspore/dataset/vision/py_transforms.py

@@ -30,7 +30,7 @@ from . import py_transforms_util as util
 from .c_transforms import parse_padding
 from .validators import check_prob, check_crop, check_resize_interpolation, check_random_resize_crop, \
     check_normalize_py, check_random_crop, check_random_color_adjust, check_random_rotation, \
-    check_transforms_list, check_random_apply, check_ten_crop, check_num_channels, check_pad, \
+    check_ten_crop, check_num_channels, check_pad, \
     check_random_perspective, check_random_erasing, check_cutout, check_linear_transform, check_random_affine, \
     check_mix_up, check_positive_degrees, check_uniform_augment_py, check_auto_contrast
 from .utils import Inter, Border
@@ -609,107 +609,6 @@ class RandomRotation:
         return util.random_rotation(img, self.degrees, self.resample, self.expand, self.center, self.fill_value)
 
 
-class RandomOrder:
-    """
-    Perform a series of transforms to the input PIL image in a random order.
-
-    Args:
-        transforms (list): List of the transformations to apply.
-
-    Examples:
-        >>> import mindspore.dataset.vision.py_transforms as py_vision
-        >>> from mindspore.dataset.transforms.py_transforms import Compose
-        >>>
-        >>> Compose([py_vision.Decode(),
-        >>>          py_vision.RandomOrder(transforms_list),
-        >>>          py_vision.ToTensor()])
-    """
-
-    @check_transforms_list
-    def __init__(self, transforms):
-        self.transforms = transforms
-
-    def __call__(self, img):
-        """
-        Call method.
-
-        Args:
-            img (PIL image): Image to apply transformations in a random order.
-
-        Returns:
-            img (PIL image), Transformed image.
-        """
-        return util.random_order(img, self.transforms)
-
-
-class RandomApply:
-    """
-    Randomly perform a series of transforms with a given probability.
-
-    Args:
-        transforms (list): List of transformations to apply.
-        prob (float, optional): The probability to apply the transformation list (default=0.5).
-
-    Examples:
-        >>> import mindspore.dataset.vision.py_transforms as py_vision
-        >>> from mindspore.dataset.transforms.py_transforms import Compose
-        >>>
-        >>> Compose([py_vision.Decode(),
-        >>>          py_vision.RandomApply(transforms_list, prob=0.6),
-        >>>          py_vision.ToTensor()])
-    """
-
-    @check_random_apply
-    def __init__(self, transforms, prob=0.5):
-        self.prob = prob
-        self.transforms = transforms
-
-    def __call__(self, img):
-        """
-        Call method.
-
-        Args:
-            img (PIL image): Image to be randomly applied a list transformations.
-
-        Returns:
-            img (PIL image), Transformed image.
-        """
-        return util.random_apply(img, self.transforms, self.prob)
-
-
-class RandomChoice:
-    """
-    Randomly select one transform from a series of transforms and apply that transform on the image.
-
-    Args:
-         transforms (list): List of transformations to be chosen from to apply.
-
-    Examples:
-        >>> import mindspore.dataset.vision.py_transforms as py_vision
-        >>> from mindspore.dataset.transforms.py_transforms import Compose
-        >>>
-        >>> Compose([py_vision.Decode(),
-        >>>          py_vision.RandomChoice(transforms_list),
-        >>>          py_vision.ToTensor()])
-    """
-
-    @check_transforms_list
-    def __init__(self, transforms):
-        self.transforms = transforms
-
-    def __call__(self, img):
-        """
-        Call method.
-
-        Args:
-            img (PIL image): Image to apply transformation.
-
-        Returns:
-            img (PIL image), Transformed image.
-        """
-        return util.random_choice(img, self.transforms)
-
-
 class FiveCrop:
     """
     Generate 5 cropped images (one central image and four corners images).

mindspore/dataset/vision/py_transforms_util.py

@@ -701,56 +701,6 @@ def random_rotation(img, degrees, resample, expand, center, fill_value):
     return rotate(img, angle, resample, expand, center, fill_value)
 
 
-def random_order(img, transforms):
-    """
-    Applies a list of transforms in a random order.
-
-    Args:
-        img: Image to be applied transformations in a random order.
-        transforms (list): List of the transformations to be applied.
-
-    Returns:
-        img, Transformed image.
-    """
-    random.shuffle(transforms)
-    for transform in transforms:
-        img = transform(img)
-    return img
-
-
-def random_apply(img, transforms, prob):
-    """
-    Apply a list of transformation, randomly with a given probability.
-
-    Args:
-        img: Image to be randomly applied a list transformations.
-        transforms (list): List of transformations to be applied.
-        prob (float): The probability to apply the transformation list.
-
-    Returns:
-        img, Transformed image.
-    """
-    if prob < random.random():
-        return img
-    for transform in transforms:
-        img = transform(img)
-    return img
-
-
-def random_choice(img, transforms):
-    """
-    Random selects one transform from a list of transforms and applies that on the image.
-
-    Args:
-        img: Image to be applied transformation.
-        transforms (list): List of transformations to be chosen from to apply.
-
-    Returns:
-        img, Transformed image.
-    """
-    return random.choice(transforms)(img)
-
-
 def five_crop(img, size):
     """
     Generate 5 cropped images (one central and four corners).

mindspore/dataset/vision/validators.py

@@ -347,37 +347,6 @@ def check_random_rotation(method):
     return new_method
 
 
-def check_transforms_list(method):
-    """Wrapper method to check the parameters of transform list."""
-
-    @wraps(method)
-    def new_method(self, *args, **kwargs):
-        [transforms], _ = parse_user_args(method, *args, **kwargs)
-
-        type_check(transforms, (list,), "transforms")
-
-        return method(self, *args, **kwargs)
-
-    return new_method
-
-
-def check_random_apply(method):
-    """Wrapper method to check the parameters of random apply."""
-
-    @wraps(method)
-    def new_method(self, *args, **kwargs):
-        [transforms, prob], _ = parse_user_args(method, *args, **kwargs)
-        type_check(transforms, (list,), "transforms")
-
-        if prob is not None:
-            type_check(prob, (float, int,), "prob")
-            check_value(prob, [0., 1.], "prob")
-
-        return method(self, *args, **kwargs)
-
-    return new_method
-
-
 def check_ten_crop(method):
     """Wrapper method to check the parameters of crop."""
 
@@ -678,7 +647,6 @@ def check_positive_degrees(method):
     return new_method
 
 
-
 def check_random_select_subpolicy_op(method):
     """Wrapper method to check the parameters of RandomSelectSubpolicyOp."""
 

tests/ut/python/dataset/test_random_apply.py

@@ -17,7 +17,7 @@ Testing RandomApply op in DE
 """
 import numpy as np
 import mindspore.dataset as ds
-import mindspore.dataset.transforms.py_transforms
+import mindspore.dataset.transforms.py_transforms as py_transforms
 import mindspore.dataset.vision.py_transforms as py_vision
 from mindspore import log as logger
 from util import visualize_list, config_get_set_seed, \
@@ -38,16 +38,16 @@ def test_random_apply_op(plot=False):
     transforms_list = [py_vision.CenterCrop(64), py_vision.RandomRotation(30)]
     transforms1 = [
         py_vision.Decode(),
-        py_vision.RandomApply(transforms_list, prob=0.6),
+        py_transforms.RandomApply(transforms_list, prob=0.6),
         py_vision.ToTensor()
     ]
-    transform1 = mindspore.dataset.transforms.py_transforms.Compose(transforms1)
+    transform1 = py_transforms.Compose(transforms1)
 
     transforms2 = [
         py_vision.Decode(),
         py_vision.ToTensor()
     ]
-    transform2 = mindspore.dataset.transforms.py_transforms.Compose(transforms2)
+    transform2 = py_transforms.Compose(transforms2)
 
     #  First dataset
     data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
@@ -79,10 +79,10 @@ def test_random_apply_md5():
     transforms = [
         py_vision.Decode(),
         # Note: using default value "prob=0.5"
-        py_vision.RandomApply(transforms_list),
+        py_transforms.RandomApply(transforms_list),
         py_vision.ToTensor()
     ]
-    transform = mindspore.dataset.transforms.py_transforms.Compose(transforms)
+    transform = py_transforms.Compose(transforms)
 
     #  Generate dataset
     data = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
@@ -111,10 +111,10 @@ def test_random_apply_exception_random_crop_badinput():
                        py_vision.RandomRotation(30)]
     transforms = [
         py_vision.Decode(),
-        py_vision.RandomApply(transforms_list, prob=0.6),
+        py_transforms.RandomApply(transforms_list, prob=0.6),
         py_vision.ToTensor()
     ]
-    transform = mindspore.dataset.transforms.py_transforms.Compose(transforms)
+    transform = py_transforms.Compose(transforms)
     #  Generate dataset
     data = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
     data = data.map(input_columns=["image"], operations=transform)

tests/ut/python/dataset/test_random_choice.py

@@ -17,7 +17,7 @@ Testing RandomChoice op in DE
 """
 import numpy as np
 import mindspore.dataset as ds
-import mindspore.dataset.transforms.py_transforms
+import mindspore.dataset.transforms.py_transforms as py_transforms
 import mindspore.dataset.vision.py_transforms as py_vision
 from mindspore import log as logger
 from util import visualize_list, diff_mse
@@ -35,16 +35,16 @@ def test_random_choice_op(plot=False):
     transforms_list = [py_vision.CenterCrop(64), py_vision.RandomRotation(30)]
     transforms1 = [
         py_vision.Decode(),
-        py_vision.RandomChoice(transforms_list),
+        py_transforms.RandomChoice(transforms_list),
         py_vision.ToTensor()
     ]
-    transform1 = mindspore.dataset.transforms.py_transforms.Compose(transforms1)
+    transform1 = py_transforms.Compose(transforms1)
 
     transforms2 = [
         py_vision.Decode(),
         py_vision.ToTensor()
     ]
-    transform2 = mindspore.dataset.transforms.py_transforms.Compose(transforms2)
+    transform2 = py_transforms.Compose(transforms2)
 
     #  First dataset
     data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
@@ -73,17 +73,17 @@ def test_random_choice_comp(plot=False):
     transforms_list = [py_vision.CenterCrop(64)]
     transforms1 = [
         py_vision.Decode(),
-        py_vision.RandomChoice(transforms_list),
+        py_transforms.RandomChoice(transforms_list),
         py_vision.ToTensor()
     ]
-    transform1 = mindspore.dataset.transforms.py_transforms.Compose(transforms1)
+    transform1 = py_transforms.Compose(transforms1)
 
     transforms2 = [
         py_vision.Decode(),
         py_vision.CenterCrop(64),
         py_vision.ToTensor()
     ]
-    transform2 = mindspore.dataset.transforms.py_transforms.Compose(transforms2)
+    transform2 = py_transforms.Compose(transforms2)
 
     #  First dataset
     data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
@@ -117,10 +117,10 @@ def test_random_choice_exception_random_crop_badinput():
     transforms_list = [py_vision.RandomCrop(5000)]
     transforms = [
         py_vision.Decode(),
-        py_vision.RandomChoice(transforms_list),
+        py_transforms.RandomChoice(transforms_list),
         py_vision.ToTensor()
     ]
-    transform = mindspore.dataset.transforms.py_transforms.Compose(transforms)
+    transform = py_transforms.Compose(transforms)
     #  Generate dataset
     data = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
     data = data.map(input_columns=["image"], operations=transform)

tests/ut/python/dataset/test_random_order.py

@@ -17,7 +17,7 @@ Testing RandomOrder op in DE
 """
 import numpy as np
 import mindspore.dataset as ds
-import mindspore.dataset.transforms.py_transforms
+import mindspore.dataset.transforms.py_transforms as py_transforms
 import mindspore.dataset.vision.py_transforms as py_vision
 from mindspore import log as logger
 from util import visualize_list, config_get_set_seed, \
@@ -38,16 +38,16 @@ def test_random_order_op(plot=False):
     transforms_list = [py_vision.CenterCrop(64), py_vision.RandomRotation(30)]
     transforms1 = [
         py_vision.Decode(),
-        py_vision.RandomOrder(transforms_list),
+        py_transforms.RandomOrder(transforms_list),
         py_vision.ToTensor()
     ]
-    transform1 = mindspore.dataset.transforms.py_transforms.Compose(transforms1)
+    transform1 = py_transforms.Compose(transforms1)
 
     transforms2 = [
         py_vision.Decode(),
         py_vision.ToTensor()
     ]
-    transform2 = mindspore.dataset.transforms.py_transforms.Compose(transforms2)
+    transform2 = py_transforms.Compose(transforms2)
 
     #  First dataset
     data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
@@ -78,10 +78,10 @@ def test_random_order_md5():
     transforms_list = [py_vision.RandomCrop(64), py_vision.RandomRotation(30)]
     transforms = [
         py_vision.Decode(),
-        py_vision.RandomOrder(transforms_list),
+        py_transforms.RandomOrder(transforms_list),
         py_vision.ToTensor()
     ]
-    transform = mindspore.dataset.transforms.py_transforms.Compose(transforms)
+    transform = py_transforms.Compose(transforms)
 
     #  Generate dataset
     data = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)