add tensor.where

docs/api/api_python/mindspore.ops.rst

@@ -442,6 +442,7 @@ Array操作
     mindspore.ops.unsorted_segment_sum
     mindspore.ops.unsqueeze
     mindspore.ops.unstack
+    mindspore.ops.where
 
 类型转换
 ^^^^^^^^^^^^^^^^

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.where.rst

@@ -0,0 +1,6 @@
+mindspore.Tensor.where
+=======================
+
+.. py:method:: mindspore.Tensor.where(condition, y)
+
+    详情请参考 :func:`mindspore.ops.where`。

docs/api/api_python/mindspore/mindspore.Tensor.rst

@@ -250,5 +250,6 @@ mindspore.Tensor
     mindspore.Tensor.unsqueeze
     mindspore.Tensor.var
     mindspore.Tensor.view
+    mindspore.Tensor.where
     mindspore.Tensor.xdivy
     mindspore.Tensor.xlogy

docs/api/api_python/ops/mindspore.ops.func_where.rst

@@ -0,0 +1,21 @@
+mindspore.ops.where
+====================
+
+.. py:function:: mindspore.ops.where(condition, x, y)
+
+    返回一个Tensor，Tensor的元素从 `x` 或 `y` 中根据 `condition` 选择。
+
+    .. math::
+
+        output_i = \begin{cases} x_i,\quad &if\ condition_i \\ y_i,\quad &otherwise \end{cases}
+
+    参数：
+        - **condition** (Union[Bool Tensor, bool, scalar]) - 如果是True，选取 `x` 中的元素，否则选取 `y` 中的元素。
+        - **x** (Union[Tensor, Scalar]) - 在 `condition` 为True的索引处选择的值。
+        - **y** (Union[Tensor, Scalar]) - 当 `condition` 为False的索引处选择的值。
+
+    返回：
+        Tensor，其中的元素从 `x` 和 `y` 中选取。
+
+    异常：
+        -**ValueError** - `condition` 不可以被广播成 `x` 的shape。

docs/api/api_python_en/Tensor_list.rst

@@ -256,6 +256,7 @@
     mindspore.Tensor.unsqueeze
     mindspore.Tensor.var
     mindspore.Tensor.view
+    mindspore.Tensor.where
     mindspore.Tensor.xdivy
     mindspore.Tensor.xlogy
 

docs/api/api_python_en/mindspore.ops.rst

@@ -442,6 +442,7 @@ Array Operation
     mindspore.ops.unsorted_segment_sum
     mindspore.ops.unsqueeze
     mindspore.ops.unstack
+    mindspore.ops.where
 
 Type Conversion
 ^^^^^^^^^^^^^^^

mindspore/ccsrc/pipeline/jit/resource.cc

@@ -413,6 +413,7 @@ BuiltInTypeMap &GetMethodMap() {
        {"sinh", std::string("sinh")},                                      // sinh()
        {"sort", std::string("sort")},                                      // sort()
        {"trunc", std::string("trunc")},                                    // trunc()
+       {"where", std::string("where")},                                    // where()
        {"imag", std::string("imag")},                                      // imag()
      }},
     {kObjectTypeRowTensorType,

mindspore/python/mindspore/_extends/parse/standard_method.py

@@ -1001,7 +1001,8 @@ def copy(x):
     return x
 
 
-def max(x, axis=None, keepdims=False, initial=None, where=True):  # pylint: disable=redefined-builtin
+def max(x, axis=None, keepdims=False, initial=None,  # pylint: disable=redefined-builtin
+        where=True):  # pylint: disable=redefined-outer-name
     """
     Returns the maximum of a tensor or maximum along an axis.
 
@@ -1046,7 +1047,8 @@ def max(x, axis=None, keepdims=False, initial=None, where=True):  # pylint: disa
                                  axis=axis, keepdims=keepdims, initial=initial, where=where)
 
 
-def min(x, axis=None, keepdims=False, initial=None, where=True):  # pylint: disable=redefined-builtin
+def min(x, axis=None, keepdims=False, initial=None,  # pylint: disable=redefined-builtin
+        where=True):  # pylint: disable=redefined-outer-name
     """
     Returns the minimum of a tensor or minimum along an axis.
 
@@ -3977,6 +3979,14 @@ def trunc(input):
     return F.trunc(input)
 
 
+def where(x, condition, y):
+    r"""
+    Returns a tensor whose elements are selected from either `x` or `y` depending on `condition`.
+    Please refer to :func:`mindspore.ops.where`.
+    """
+    return F.where(condition, x, y)
+
+
 def imag(input):
     r"""
     Returns a new tensor containing imaginary value of the input.

mindspore/python/mindspore/common/tensor.py

@@ -4488,6 +4488,13 @@ class Tensor(Tensor_):
         self._init_check()
         return tensor_operator_registry.get('trunc')(self)
 
+    def where(self, condition, y):
+        r"""
+        For details, please refer to :func:`mindspore.ops.where`
+        """
+        self._init_check()
+        return tensor_operator_registry.get('where')(condition, self, y)
+
     def imag(self):
         r"""
         Returns a new tensor containing imaginary value of the input tensor.

mindspore/python/mindspore/numpy/array_ops.py

@@ -711,8 +711,7 @@ def where(condition, x=None, y=None):
         x = F.cast(x, dtype)
     if not _check_same_type(dtype2, dtype):
         y = F.cast(y, dtype)
-    is_bool = _check_same_type(dtype1, mstype.bool_) and _check_same_type(
-        dtype2, mstype.bool_)
+    is_bool = _check_same_type(dtype1, mstype.bool_) and _check_same_type(dtype2, mstype.bool_)
     if is_bool:
         # select does not support bool type for x or y
         x = F.cast(x, mstype.float32)

mindspore/python/mindspore/ops/function/__init__.py

@@ -105,6 +105,7 @@ from .array_func import (
     matrix_set_diag,
     diag,
     masked_select,
+    where,
     meshgrid,
     affine_grid,
     fills,

mindspore/python/mindspore/ops/function/array_func.py

@@ -268,6 +268,40 @@ def eye(n, m, t):
     return eye_(n, m, t)
 
 
+def where(condition, x, y):
+    r"""
+    Returns a tensor whose elements are selected from either `x` or `y` depending on `condition`.
+
+    ..math::
+        output_i = \begin{cases} x_i,\quad &if\ condition_i \\ y_i,\quad &otherwise \end{cases}
+
+    Args:
+        condition (Bool Tensor, bool, scalar): If True, yield `x` otherwise yield `y`.
+        x (Union[Tensor, Scalar]): Value (if `x` is a scalar) or values selected at indices where condition is True.
+        y (Union[Tensor, Scalar]): Value (if `y` is a scalar) or values selected at indices where condition is False.
+
+    Returns:
+        Tensor, elements are selected from `x` and `y`.
+
+    Raises:
+        ValueError: If `condition` can not be broadcast to the shape of `x`.
+
+    Examples:
+        >>> a = Tensor(np.arange(4).reshape((2, 2)), mstype.float32)
+        >>> b = Tensor(np.ones((2, 2)), mstype.float16)
+        >>> condition = a < 3
+        >>> output = ops.where(condition, a, b)
+        >>> print(output)
+        [[0. 1.]
+         [2. 1.]]
+    """
+    not_op = P.LogicalNot()
+    x_mask = cast_(condition, mstype.bool_)
+    y_mask = not_op(x_mask)
+    output = x * x_mask + y * y_mask
+    return output
+
+
 def reverse(x, axis):
     """
     Reverses specific dimensions of a tensor.
@@ -5226,6 +5260,7 @@ __all__ = [
     'one_hot',
     'masked_fill',
     'masked_select',
+    'where',
     'narrow',
     'scatter_add',
     'scatter_mul',

mindspore/python/mindspore/ops/function/math_func.py

@@ -712,7 +712,8 @@ def subtract(x, other, *, alpha=1):
         output[i] = x[i] - alpha * y[i]
 
     Args:
-        other (Union[Tensor, number.Number]): The tensor or number to be subtracted.
+        x (Union[Tensor, number.Number]): The tensor or number to be subtracted.
+        other (Union[Tensor, number.Number]): The tensor or number to subtract.
 
     Keyword Args:
         alpha (Number): The multiplier for `other`. Default: 1.

mindspore/python/mindspore/ops/functional.py

@@ -356,6 +356,7 @@ tensor_operator_registry.register('ne', ne)
 tensor_operator_registry.register('sinh', sinh)
 tensor_operator_registry.register('sort', P.Sort)
 tensor_operator_registry.register('trunc', trunc)
+tensor_operator_registry.register('where', where)
 tensor_operator_registry.register('imag', imag)
 tensor_operator_registry.register('repeat_interleave', repeat_interleave)
 tensor_operator_registry.register('rad2deg', rad2deg)

tests/st/ops/test_ops_where.py

@@ -0,0 +1,36 @@
+import numpy as np
+import pytest
+import mindspore.common.dtype as mstype
+import mindspore.nn as nn
+import mindspore.ops as ops
+from mindspore import Tensor
+from mindspore import context
+
+
+class Net(nn.Cell):
+    def construct(self, condition, x, y):
+        return ops.where(condition, x, y)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_cpu
+@pytest.mark.platform_arm_cpu
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.platform_arm_ascend_training
+@pytest.mark.platform_x86_ascend_training
+@pytest.mark.env_onecard
+@pytest.mark.parametrize('mode', [context.GRAPH_MODE, context.PYNATIVE_MODE])
+def test_ops_where(mode):
+    """
+    Feature: ops.where
+    Description: Verify the result of ops.where
+    Expectation: success
+    """
+    context.set_context(mode=mode)
+    net = Net()
+    x = Tensor(np.arange(4).reshape((2, 2)), mstype.float32)
+    y = Tensor(np.ones((2, 2)), mstype.float32)
+    condition = x < 3
+    output = net(condition, x, y)
+    expected = np.array([[0, 1], [2, 1]], dtype=np.float32)
+    assert np.allclose(output.asnumpy(), expected)

tests/st/tensor/test_where.py

@@ -0,0 +1,47 @@
+# Copyright 2022 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.
+# ============================================================================
+import numpy as np
+import pytest
+import mindspore.common.dtype as mstype
+import mindspore.nn as nn
+from mindspore import Tensor
+from mindspore import context
+
+
+class Net(nn.Cell):
+    def construct(self, condition, x, y):
+        return x.where(condition, y)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.platform_arm_ascend_training
+@pytest.mark.platform_x86_ascend_training
+@pytest.mark.env_onecard
+@pytest.mark.parametrize('mode', [context.GRAPH_MODE, context.PYNATIVE_MODE])
+def test_tensor_where(mode):
+    """
+    Feature: tensor.where
+    Description: Verify the result of where
+    Expectation: success
+    """
+    context.set_context(mode=mode)
+    x = Tensor(np.arange(4).reshape((2, 2)), mstype.float32)
+    y = Tensor(np.ones((2, 2)), mstype.float32)
+    condition = x < 3
+    net = Net()
+    output = net(condition, x, y)
+    expected = np.array([[0, 1], [2, 1]], dtype=np.float32)
+    assert np.allclose(output.asnumpy(), expected)