tensor_inverse_master

2022-11-22 11:14:49 +08:00 · 2022-11-22 11:14:49 +08:00 · c4b0223330
parent dc5d0d782c
commit c4b0223330
16 changed files with 175 additions and 7 deletions
--- a/docs/api/api_python/mindspore.ops.rst
+++ b/docs/api/api_python/mindspore.ops.rst
@ -203,6 +203,7 @@ mindspore.ops
    mindspore.ops.hypot
    mindspore.ops.i0
    mindspore.ops.inv
+    mindspore.ops.inverse
    mindspore.ops.invert
    mindspore.ops.lcm
    mindspore.ops.ldexp
--- a/docs/api/api_python/mindspore/Tensor/mindspore.Tensor.inverse.rst
+++ b/docs/api/api_python/mindspore/Tensor/mindspore.Tensor.inverse.rst
@ -0,0 +1,6 @@
+mindspore.Tensor.inverse
+========================
+
+.. py:method:: mindspore.Tensor.inverse()
+
+    详情请参考 :func:`mindspore.ops.inverse`。
--- a/docs/api/api_python/mindspore/mindspore.Tensor.rst
+++ b/docs/api/api_python/mindspore/mindspore.Tensor.rst
@ -129,6 +129,7 @@ mindspore.Tensor
    mindspore.Tensor.inplace_update
    mindspore.Tensor.int
    mindspore.Tensor.inv
+    mindspore.Tensor.inverse
    mindspore.Tensor.invert
    mindspore.Tensor.isclose
    mindspore.Tensor.isfinite
--- a/docs/api/api_python/ops/mindspore.ops.func_inverse.rst
+++ b/docs/api/api_python/ops/mindspore.ops.func_inverse.rst
@ -0,0 +1,17 @@
+mindspore.ops.inverse
+=====================
+
+.. py:function:: mindspore.ops.inverse(x)
+
+    计算输入矩阵的逆。
+
+    参数：
+        - **x** (Tensor) - 计算的矩阵。`x` 至少是两维的，最后两个维度大小相同。
+
+    返回：
+        Tensor，shape和类型和 `x` 相同。
+
+    异常：
+        - **TypeError** - `x` 不是Tensor。
+        - **ValueError** - `x` 最后两个维度的大小不相同。
+        - **ValueError** - `x` 的维数小于2。
--- a/docs/api/api_python_en/Tensor_list.rst
+++ b/docs/api/api_python_en/Tensor_list.rst
@ -135,6 +135,7 @@
    mindspore.Tensor.inplace_update
    mindspore.Tensor.int
    mindspore.Tensor.inv
+    mindspore.Tensor.inverse
    mindspore.Tensor.invert
    mindspore.Tensor.isclose
    mindspore.Tensor.isfinite
--- a/docs/api/api_python_en/mindspore.ops.rst
+++ b/docs/api/api_python_en/mindspore.ops.rst
@ -204,6 +204,7 @@ Element-by-Element Operations
    mindspore.ops.hypot
    mindspore.ops.i0
    mindspore.ops.inv
+    mindspore.ops.inverse
    mindspore.ops.invert
    mindspore.ops.lcm
    mindspore.ops.ldexp
--- a/mindspore/ccsrc/pipeline/jit/resource.cc
+++ b/mindspore/ccsrc/pipeline/jit/resource.cc
@ -287,6 +287,7 @@ BuiltInTypeMap &GetMethodMap() {
       {"is_floating_point", std::string("is_floating_point")},            // is_floating_point()
       {"is_signed", std::string("is_signed")},                            // is_signed()
       {"inv", std::string("inv")},                                        // inv()
+       {"inverse", std::string("inverse")},                                // inverse()
       {"invert", std::string("invert")},                                  // invert()
       {"searchsorted", std::string("searchsorted")},                      // P.Select()
       {"take", std::string("take")},                                      // P.GatherNd()
--- a/mindspore/python/mindspore/_extends/parse/standard_method.py
+++ b/mindspore/python/mindspore/_extends/parse/standard_method.py
@ -1418,6 +1418,13 @@ def inv(x):
    return F.inv(x)


+def inverse(x):
+    """
+    Computes the inverse of a square matrix.
+    """
+    return F.inverse(x)
+
+
 def invert(x):
    """
    Flips all bits of input tensor element-wise.
--- a/mindspore/python/mindspore/common/tensor.py
+++ b/mindspore/python/mindspore/common/tensor.py
@ -1478,6 +1478,13 @@ class Tensor(Tensor_):
        self._init_check()
        return tensor_operator_registry.get('inv')(self)

+    def inverse(self):
+        r"""
+        For details, please refer to :func:`mindspore.ops.inverse`.
+        """
+        self._init_check()
+        return tensor_operator_registry.get('inverse')(self)
+
    def invert(self):
        r"""
        For details, please refer to :func:`mindspore.ops.invert`.
--- a/mindspore/python/mindspore/ops/function/init.py
+++ b/mindspore/python/mindspore/ops/function/init.py
@ -216,6 +216,7 @@ from .math_func import (
    inplace_sub,
    inplace_update,
    inv,
+    inverse,
    invert,
    minimum,
    renorm,
--- a/mindspore/python/mindspore/ops/function/math_func.py
+++ b/mindspore/python/mindspore/ops/function/math_func.py
@ -2123,6 +2123,36 @@ def inv(x):
    return inv_(x)


+def inverse(x):
+    """
+    Compute the inverse of the input matrix.
+
+    Inputs:
+        - **x** (Tensor) - A matrix to be calculated. Input `x` must be at least two dimensions, and the size of
+          the last two dimensions must be the same size.
+
+    Returns:
+        Tensor, has the same type and shape as input `x`.
+
+    Raises:
+        TypeError: If `x` is not a Tensor.
+        ValueError: If the size of the last two dimensions of `x` is not the same.
+        ValueError: If the dimension of `x` is less than 2.
+
+    Supported Platforms:
+        ``GPU`` ``CPU``
+
+    Examples:
+        >>> x = Tensor([[1., 2.], [3., 4.]], ms.float32)
+        >>> print(ops.inverse(x))
+        [[-2.   1. ]
+         [ 1.5 -0.5]]
+    """
+    if x.dtype in mstype.int_type:
+        _get_cache_prim(P.Cast)()(x, mstype.float64)
+    return _get_cache_prim(P.MatrixInverse)()(x)
+
+
 def invert(x):
    r"""
    Flips all bits of input tensor element-wise.
@ -7484,6 +7514,7 @@ __all__ = [
    'bitwise_or',
    'bitwise_xor',
    'inv',
+    'inverse',
    'invert',
    'erf',
    'erfc',
--- a/mindspore/python/mindspore/ops/function/nn_func.py
+++ b/mindspore/python/mindspore/ops/function/nn_func.py
@ -1419,7 +1419,7 @@ def fractional_max_pool2d(input_x, kernel_size, output_size=None, output_ratio=N
        raise ValueError(f"For fractional_max_pool2d, 'output_size' and 'output_ratio' can not be specified or None"
                         f"at the same time, but got {output_ratio} and {output_size} .")
    if len(input_x.shape) == 3:
-        input_x.expend_dims(axis=0)
+        input_x = input_x.expand_dims(axis=0)
    if _random_samples is None:
        _random_samples = Tensor([[[0, 0]]], mstype.float32)
    if output_ratio is not None:
@ -1517,8 +1517,6 @@ def fractional_max_pool3d(input_x, kernel_size, output_size=None, output_ratio=N
    if output_ratio is not None and output_size is not None or output_ratio is None and output_size is None:
        raise ValueError(f"For fractional_max_pool2d, 'output_size' and 'output_ratio' can not be specified or None"
                         f"at the same time, but got {output_ratio} and {output_size} .")
-    if len(input_x.shape) == 4:
-        input_x.expend_dims(axis=0)
    if _random_samples is None:
        _random_samples = Tensor([[[0, 0, 0]]], mstype.float32)
    if output_ratio is not None:
--- a/mindspore/python/mindspore/ops/functional.py
+++ b/mindspore/python/mindspore/ops/functional.py
@ -204,6 +204,7 @@ tensor_operator_registry.register('inv', inv)
 tensor_operator_registry.register('logaddexp', logaddexp)
 tensor_operator_registry.register('logaddexp2', logaddexp2)
 tensor_operator_registry.register('logsumexp', logsumexp)
+tensor_operator_registry.register('inverse', inverse)
 tensor_operator_registry.register('invert', invert)
 tensor_operator_registry.register('hardshrink', P.HShrink)
 tensor_operator_registry.register('heaviside', heaviside)
--- a/tests/st/ops/test_ops_fractional_max_pool.py
+++ b/tests/st/ops/test_ops_fractional_max_pool.py
@ -44,7 +44,7 @@ def test_fractional_maxpool2d_normal(mode):
    """
    ms.set_context(mode=mode)
    net = FractionalMaxPool2dNet()
-    input_x = Tensor(np.random.rand(25).reshape([1, 1, 5, 5]), mstype.float32)
+    input_x = Tensor(np.random.rand(25).reshape([1, 5, 5]), mstype.float32)
    output1, output2 = net(input_x)
    assert output1[0].shape == output1[1].shape == (1, 1, 2, 2)
    assert output2[0].shape == output2[1].shape == (1, 1, 2, 2)
@ -88,11 +88,11 @@ def test_fractional_maxpool3d_normal(mode):
    Expectation: Success
    """
    ms.set_context(mode=mode)
-    input_x = Tensor(np.random.rand(16).reshape([1, 1, 2, 2, 4]), mstype.float32)
+    input_x = Tensor(np.random.rand(16).reshape([1, 2, 2, 4]), mstype.float32)
    net = FractionalMaxPool3dNet()
    output1, output2 = net(input_x)
-    assert output1[0].shape == output1[1].shape == (1, 1, 1, 1, 2)
-    assert output2[0].shape == output2[1].shape == (1, 1, 1, 1, 2)
+    assert output1[0].shape == output1[1].shape == (1, 1, 1, 2)
+    assert output2[0].shape == output2[1].shape == (1, 1, 1, 2)
    input_x = Tensor([[[[[5.76273143e-001, 7.97047436e-001, 5.05385816e-001, 7.98332036e-001],
                         [5.79880655e-001, 9.75979388e-001, 3.17571498e-002, 8.08261558e-002]],
                        [[3.82758647e-001, 7.09801614e-001, 4.39641386e-001, 5.71077049e-001],
--- a/tests/st/ops/test_ops_inverse.py
+++ b/tests/st/ops/test_ops_inverse.py
@ -0,0 +1,48 @@
+# 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 as ms
+import mindspore.nn as nn
+from mindspore import Tensor, ops
+
+
+class Net(nn.Cell):
+    def construct(self, x):
+        return ops.inverse(x)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_cpu
+@pytest.mark.platform_arm_cpu
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+@pytest.mark.parametrize('mode', [ms.GRAPH_MODE, ms.PYNATIVE_MODE])
+def test_ops_inverse(mode):
+    """
+    Feature: ops.inverse
+    Description: Verify the result of inverse
+    Expectation: success
+    """
+    ms.set_context(mode=mode)
+    x = Tensor([[1., 2, 3],
+                [4, 5., 6],
+                [8, 8, 9]], ms.float32)
+    net = Net()
+    output = net(x)
+    expect_output = [[1.0000008, -2.000001, 1.0000005],
+                     [-4.0000014, 5.000002, -2.000001],
+                     [2.6666675, -2.6666675, 1.0000002]]
+    assert np.allclose(output.asnumpy(), expect_output)
--- a/tests/st/tensor/test_inverse.py
+++ b/tests/st/tensor/test_inverse.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 as ms
+import mindspore.nn as nn
+
+
+class Net(nn.Cell):
+    def construct(self, x):
+        return x.inverse()
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_cpu
+@pytest.mark.platform_arm_cpu
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+@pytest.mark.parametrize('mode', [ms.GRAPH_MODE, ms.PYNATIVE_MODE])
+def test_tensor_inverse(mode):
+    """
+    Feature: tensor.inverse
+    Description: Verify the result of inverse
+    Expectation: success
+    """
+    ms.set_context(mode=mode)
+    x = ms.Tensor([[1., 2, 3],
+                   [4, 5., 6],
+                   [8, 8, 9]], ms.float32)
+    net = Net()
+    output = net(x)
+    expect_output = [[1.0000008, -2.000001, 1.0000005],
+                     [-4.0000014, 5.000002, -2.000001],
+                     [2.6666675, -2.6666675, 1.0000002]]
+    assert np.allclose(output.asnumpy(), expect_output)