【PaddlePaddle Hackathon 3 No.20】为 Paddle 新增 vsplit API (#44853)

* add paddle vsplit api

* update unittest and fix a typo

* update

* add vsplit to __all__

* update unit test and description of x

* fix typo

python/paddle/__init__.py

@@ -165,6 +165,7 @@ from .tensor.manipulation import shard_index  # noqa: F401
 from .tensor.manipulation import slice  # noqa: F401
 from .tensor.manipulation import crop  # noqa: F401
 from .tensor.manipulation import split  # noqa: F401
+from .tensor.manipulation import vsplit  # noqa: F401
 from .tensor.manipulation import squeeze  # noqa: F401
 from .tensor.manipulation import squeeze_  # noqa: F401
 from .tensor.manipulation import stack  # noqa: F401
@@ -457,6 +458,7 @@ __all__ = [  # noqa
     'searchsorted',
     'bucketize',
     'split',
+    'vsplit',
     'logical_and',
     'full_like',
     'less_than',

python/paddle/fluid/tests/unittests/test_splits_api.py

+# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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.
+
+import unittest
+import numpy as np
+import paddle
+import paddle.fluid as fluid
+import paddle.fluid.core as core
+
+
+def func_ref(func, x, num_or_sections):
+    # Convert the num_or_sections in paddle to indices_or_sections in numpy
+    # Do not support -1
+    if isinstance(num_or_sections, int):
+        indices_or_sections = num_or_sections
+    else:
+        indices_or_sections = np.cumsum(num_or_sections)[:-1]
+    return func(x, indices_or_sections)
+
+
+# TODO: add other split API, such as dsplit、hsplit
+test_list = [
+    (paddle.vsplit, np.vsplit),
+]
+
+
+class TestSplitsAPI(unittest.TestCase):
+
+    def setUp(self):
+        self.rtol = 1e-5
+        self.atol = 1e-8
+        self.set_input()
+
+    def set_input(self):
+        self.shape = [4, 5, 2]
+        self.num_or_sections = 2
+        self.x_np = np.random.uniform(-1, 1, self.shape).astype('float64')
+        self.place = paddle.CUDAPlace(0) if core.is_compiled_with_cuda() \
+            else paddle.CPUPlace()
+
+    def test_static_api(self):
+        paddle.enable_static()
+        for func, func_type in test_list:
+            with paddle.static.program_guard(paddle.static.Program()):
+                x = paddle.fluid.data('X', self.x_np.shape, self.x_np.dtype)
+                out = func(x, self.num_or_sections)
+                exe = paddle.static.Executor(self.place)
+                res = exe.run(feed={'X': self.x_np}, fetch_list=[out])
+            out_ref = func_ref(func_type, self.x_np, self.num_or_sections)
+            for n, p in zip(out_ref, res):
+                np.testing.assert_allclose(n, p, rtol=self.rtol, atol=self.atol)
+
+    def test_dygraph_api(self):
+        paddle.disable_static(self.place)
+        x = paddle.to_tensor(self.x_np)
+        for func, func_type in test_list:
+            out = func(x, self.num_or_sections)
+            out_ref = func_ref(func_type, self.x_np, self.num_or_sections)
+            for n, p in zip(out_ref, out):
+                np.testing.assert_allclose(n,
+                                           p.numpy(),
+                                           rtol=self.rtol,
+                                           atol=self.atol)
+        paddle.enable_static()
+
+
+class TestSplitsSections(TestSplitsAPI):
+    """
+        Test num_or_sections which is a list and date type is float64.
+    """
+
+    def set_input(self):
+        self.shape = [6, 2, 4]
+        self.num_or_sections = [2, 1, 3]
+        self.x_np = np.random.uniform(-1, 1, self.shape).astype('float64')
+        self.place = paddle.CUDAPlace(0) if core.is_compiled_with_cuda() \
+            else paddle.CPUPlace()
+
+
+class TestSplitsFloat32(TestSplitsAPI):
+    """
+        Test num_or_sections which is an integer and data type is float32.
+    """
+
+    def set_input(self):
+        self.shape = [2, 3, 4]
+        self.num_or_sections = 2
+        self.x_np = np.random.uniform(-1, 1, self.shape).astype('float32')
+        self.place = paddle.CUDAPlace(0) if core.is_compiled_with_cuda() \
+            else paddle.CPUPlace()
+
+
+class TestSplitsInt32(TestSplitsAPI):
+    """
+        Test data type int32.
+    """
+
+    def set_input(self):
+        self.shape = [5, 1, 2]
+        self.num_or_sections = 5
+        self.x_np = np.random.uniform(-1, 1, self.shape).astype('int32')
+        self.place = paddle.CUDAPlace(0) if core.is_compiled_with_cuda() \
+            else paddle.CPUPlace()
+
+
+class TestSplitsInt64(TestSplitsAPI):
+    """
+        Test data type int64.
+    """
+
+    def set_input(self):
+        self.shape = [4, 3, 2]
+        self.num_or_sections = 2
+        self.x_np = np.random.uniform(-1, 1, self.shape).astype('int64')
+        self.place = paddle.CUDAPlace(0) if core.is_compiled_with_cuda() \
+            else paddle.CPUPlace()
+
+
+class TestSplitsCPU(TestSplitsAPI):
+    """
+        Test cpu place and num_or_sections which is a tuple.
+    """
+
+    def set_input(self):
+        self.shape = [8, 2, 3, 5]
+        self.num_or_sections = (2, 3, 3)
+        self.x_np = np.random.uniform(-1, 1, self.shape).astype('float64')
+        self.place = paddle.CPUPlace()
+
+
+class TestSplitsError(unittest.TestCase):
+    """
+        Test the situation that input shape less than 2.
+    """
+
+    def setUp(self):
+        self.num_or_sections = 1
+        self.place = paddle.CUDAPlace(0) if core.is_compiled_with_cuda() \
+            else paddle.CPUPlace()
+
+    def test_static_error(self):
+        paddle.enable_static()
+        for func, _ in test_list:
+            with paddle.static.program_guard(paddle.static.Program()):
+                x = paddle.fluid.data('X', [5], 'float32')
+                self.assertRaises(ValueError, func, x, self.num_or_sections)
+
+    def test_dygraph_error(self):
+        paddle.disable_static(self.place)
+        for func, _ in test_list:
+            x_np = np.random.randn(2)
+            x = paddle.to_tensor(x_np, dtype='float64')
+            self.assertRaises(ValueError, func, x, self.num_or_sections)
+
+
+if __name__ == '__main__':
+    unittest.main()

python/paddle/tensor/__init__.py

@@ -106,6 +106,7 @@ from .manipulation import scatter_nd  # noqa: F401
 from .manipulation import shard_index  # noqa: F401
 from .manipulation import slice  # noqa: F401
 from .manipulation import split  # noqa: F401
+from .manipulation import vsplit  # noqa: F401
 from .manipulation import squeeze  # noqa: F401
 from .manipulation import squeeze_  # noqa: F401
 from .manipulation import stack  # noqa: F401
@@ -428,6 +429,7 @@ tensor_method_func = [  # noqa
     'shard_index',
     'slice',
     'split',
+    'vsplit',
     'chunk',
     'tensordot',
     'squeeze',

python/paddle/tensor/manipulation.py

@@ -1926,6 +1926,48 @@ def split(x, num_or_sections, axis=0, name=None):
     return outs
 
 
+def vsplit(x, num_or_sections, name=None):
+    """
+    Split the input tensor into multiple sub-Tensors along the vertical axis, which is equivalent to ``paddle.split`` with ``axis=0``.
+    
+    Args:
+        x (Tensor): A Tensor whose dimension must be greater than 1. The data type is bool, float16, float32, float64, uint8, int8, int32 or int64.
+        num_or_sections (int|list|tuple): If ``num_or_sections`` is an int, then ``num_or_sections`` 
+            indicates the number of equal sized sub-Tensors that the ``x`` will be divided into.
+            If ``num_or_sections`` is a list or tuple, the length of it indicates the number of
+            sub-Tensors and the elements in it indicate the sizes of sub-Tensors'  dimension orderly.
+            The length of the list must not  be larger than the ``x`` 's size of axis 0.
+        name (str, optional): The default value is None.  Normally there is no need for user to set this property.
+            For more information, please refer to :ref:`api_guide_Name` .
+    Returns:
+        list[Tensor], The list of segmented Tensors.
+    
+    Example:
+        .. code-block:: python
+            
+            import paddle
+            
+            # x is a Tensor of shape [8, 6, 7]
+            x = paddle.rand([8, 6, 7])
+            out0, out1, out2 = paddle.vsplit(x, num_or_sections=2)
+            print(out0.shape)  # [4, 6, 7]
+            print(out1.shape)  # [4, 6, 7]
+            out0, out1, out2 = paddle.vsplit(x, num_or_sections=[1, 3, 4])
+            print(out0.shape)  # [1, 6, 7]
+            print(out1.shape)  # [3, 6, 7]
+            print(out2.shape)  # [4, 6, 7]
+            out0, out1, out2 = paddle.vsplit(x, num_or_sections=[2, 3, -1])
+            print(out0.shape)  # [2, 6, 7]
+            print(out1.shape)  # [3, 6, 7]
+            print(out2.shape)  # [3, 6, 7]
+    """
+    if x.ndim < 2:
+        raise ValueError(
+            "The input tensor's dimension must be greater than 1, but got {}".
+            format(x.ndim))
+    return split(x, num_or_sections, axis=0, name=name)
+
+
 def squeeze(x, axis=None, name=None):
     """
     Squeeze the dimension(s) of size 1 of input tensor x's shape.