m53297601
/
Paddle
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

(Variable --->Tensor) refine the OP doc for API2.0 (#25737)

Browse Source
fix_copy_if_different
wangchaochaohu 5 years ago committed by GitHub
parent 32b9577b2a
commit fc6fed3283
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 155 additions and 127 deletions
Show Stats Download Patch File Download Diff File

136
python/paddle/fluid/layers/tensor.py
View File

File diff suppressed because it is too large Load Diff

110
python/paddle/tensor/creation.py
Unescape View File

@ -58,25 +58,25 @@ __all__ = [
def full_like(x, fill_value, dtype=None, name=None):
"""
:alias_main: paddle.full_like
:alias: paddle.full_like,paddle.tensor.full_like,paddle.tensor.creation.full_like
:alias: paddle.tensor.full_like, paddle.tensor.creation.full_like
**full_like**
This function creates a tensor filled with `fill_value` which has identical shape and dtype
with `input`.
This function creates a tensor filled with ``fill_value`` which has identical shape of ``x`` and ``dtype``.
If the ``dtype`` is None, the data type of Tensor is same with ``x``.
Args:
x(Variable): The input tensor which specifies shape and data type. The data type can be bool, float16, float32, float64, int32, int64.
fill_value(bool|float|int|Variable): The value to fill the tensor with. Note: this value shouldn't exceed the range of the output data type.
x(Tensor): The input tensor which specifies shape and data type. The data type can be bool, float16, float32, float64, int32, int64.
fill_value(bool|float|int): The value to fill the tensor with. Note: this value shouldn't exceed the range of the output data type.
dtype(np.dtype|core.VarDesc.VarType|str, optional): The data type of output. The data type can be one
of bool, float16, float32, float64, int32, int64. The default value is None, which means the output
data type is the same as input.
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:
out(Variable): The Tensor variable storing the output.
Tensor: Tensor which is created according to ``x``, ``fill_value`` and ``dtype``.
Raises:
TypeError: The dtype must be one of bool, float16, float32, float64, int32, int64 and None.
TypeError: The data type of ``x`` must be one of bool, float16, float32, float64, int32, int64.
TypeError: The ``dtype`` must be one of bool, float16, float32, float64, int32, int64 and None.
Examples:
.. code-block:: python
@ -101,6 +101,9 @@ def full_like(x, fill_value, dtype=None, name=None):
return core.ops.fill_any_like(x, 'value', fill_value, 'dtype', dtype)
helper = LayerHelper("full_like", **locals())
check_variable_and_dtype(
x, 'x', ['bool', 'float16', 'float32', 'float64', 'int32', 'int64'],
'full_like')
check_dtype(dtype, 'dtype',
['bool', 'float16', 'float32', 'float64', 'int32', 'int64'],
'full_like/zeros_like/ones_like')
@ -119,23 +122,24 @@ def full_like(x, fill_value, dtype=None, name=None):
def ones(shape, dtype=None, name=None):
"""
:alias_main: paddle.ones
:alias: paddle.ones,paddle.tensor.ones,paddle.tensor.creation.ones
:alias: paddle.tensor.ones, paddle.tensor.creation.ones
The OP creates a tensor of specified :attr:`shape` and :attr:`dtype`, and fills it with 1.
Args:
shape(tuple|list|Variable): Shape of output tensor, the data type of shape is int32 or int64.
dtype(np.dtype|core.VarDesc.VarType|str, optional): Data type of output tensor, it supports
shape(tuple|list|Tensor): Shape of the Tensor to be created, the data type of shape is int32 or int64.
dtype(np.dtype|core.VarDesc.VarType|str, optional): Data type of output Tensor, it supports
bool, float16, float32, float64, int32 and int64. Default: if None, the data type is 'float32'.
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:
Variable: A tensor of data type :attr:`dtype` with shape :attr:`shape` and all elements set to 1.
Tensor: A tensor of data type :attr:`dtype` with shape :attr:`shape` and all elements set to 1.
Raises:
TypeError: The dtype must be one of bool, float16, float32, float64, int32, int64 and None
TypeError: The ``dtype`` must be one of bool, float16, float32, float64, int32, int64 and None
and the data type of out Tensor must be the same as the dtype.
TypeError: The `shape` must be one of list, tuple and Variable.
TypeError: The ``shape`` must be one of list, tuple and Tensor. The data type of ``shape`` must
be int32 or int64 when it's a Tensor.
Examples:
.. code-block:: python
@ -143,7 +147,7 @@ def ones(shape, dtype=None, name=None):
import paddle
paddle.enable_imperative()
#default dtype for ones OP
# default dtype for ones OP
data1 = paddle.ones(shape=[3, 2])
# [[1. 1.]
# [1. 1.]
@ -153,7 +157,7 @@ def ones(shape, dtype=None, name=None):
# [[1 1]
# [1 1]]
#shape is a Variable
# shape is a Tensor
shape = paddle.fill_constant(shape=[2], dtype='int32', value=2)
data3 = paddle.ones(shape=shape, dtype='int32')
# [[1 1]
@ -210,28 +214,45 @@ def ones_like(x, dtype=None, name=None):
def zeros(shape, dtype=None, name=None):
"""
:alias_main: paddle.zeros
:alias: paddle.zeros,paddle.tensor.zeros,paddle.tensor.creation.zeros
:alias: paddle.tensor.zeros, paddle.tensor.creation.zeros
The OP creates a tensor of specified :attr:`shape` and :attr:`dtype`, and fills it with 0.
Args:
shape(tuple|list|Variable): Shape of output tensor. The data type of shape is int32 or int64.
dtype(np.dtype|core.VarDesc.VarType|str, optional): Data type of output tensor, it supports
shape(tuple|list|Tensor): Shape of the Tensor to be created, the data type of ``shape`` is int32 or int64.
dtype(np.dtype|core.VarDesc.VarType|str, optional): Data type of output Tensor, it supports
bool, float16, float32, float64, int32 and int64. Default: if None, the date type is float32.
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:
Variable: A tensor of data type :attr:`dtype` with shape :attr:`shape` and all elements set to 0.
Tensor: A tensor of data type :attr:`dtype` with shape :attr:`shape` and all elements set to 0.
Raises:
TypeError: The ``dtype`` must be one of bool, float16, float32, float64, int32, int64 and None
and the data type of out Tensor must be the same as the dtype.
TypeError: The ``shape`` must be one of list, tuple and Tensor. The data type of ``shape`` must
be int32 or int64 when it's a Tensor.
Examples:
.. code-block:: python
import paddle
paddle.enable_imperative() # Now we are in imperative mode
data = paddle.zeros(shape=[3, 2], dtype='float32') # [[0., 0.], [0., 0.], [0., 0.]]
data = paddle.zeros(shape=[2, 2], dtype='int32', name='zeros') # [[0, 0], [0, 0]]
data = paddle.zeros(shape=[3, 2], dtype='float32')
# [[0. 0.]
# [0. 0.]
# [0. 0.]]
data = paddle.zeros(shape=[2, 2])
# [[0. 0.]
# [0. 0.]]
# shape is a Tensor
shape = paddle.fill_constant(shape=[2], dtype='int32', value=2)
data3 = paddle.ones(shape=shape, dtype='int32')
# [[0 0]
# [0 0]]
"""
if dtype is None:
dtype = 'float32'
@ -283,27 +304,31 @@ def zeros_like(x, dtype=None, name=None):
def eye(num_rows, num_columns=None, dtype=None, name=None):
"""
:alias_main: paddle.eye
:alias: paddle.tensor.eye, paddle.tensor.creation.eye
This function constructs 2-D Tensor with ones on the diagonal and zeros elsewhere.
Args:
num_rows(int): the number of rows in each batch tensor.
num_columns(int, optional): the number of columns in each batch tensor.
num_rows(int): the number of rows in each batch Tensor.
num_columns(int, optional): the number of columns in each batch Tensor.
If None, default: num_rows.
dtype(np.dtype|core.VarDesc.VarType|str, optional): The data type of the returned tensor.
dtype(np.dtype|core.VarDesc.VarType|str, optional): The data type of the returned Tensor.
It should be int32, int64, float16, float32, float64. Default: if None, the data type
is float32.
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:
Variable: An identity Tensor or LoDTensor of shape [num_rows, num_columns].
Tensor: An identity Tensor or LoDTensor of shape [num_rows, num_columns].
Raises:
TypeError: The `dtype` must be one of float16, float32, float64, int32 int64 and None.
TypeError: The `num_columns` must be non-negative int.
TypeError: The ``dtype`` must be one of float16, float32, float64, int32 int64 and None.
TypeError: The ``num_columns`` must be non-negative int.
Examples:
.. code-block:: python
import paddle
paddle.enable_imperative() # Now we are in imperative mode
@ -330,29 +355,30 @@ def eye(num_rows, num_columns=None, dtype=None, name=None):
def full(shape, fill_value, dtype=None, name=None):
"""
:alias_main: paddle.full
:alias: paddle.full,paddle.tensor.full,paddle.tensor.creation.full
:alias: paddle.tensor.full, paddle.tensor.creation.full
This Op return a Tensor with the `fill_value` which size is same as `shape`
This Op return a Tensor with the ``fill_value`` which size is same as ``shape``.
Args:
shape(list|tuple|Variable): Shape of the Tensor to be created.
shape(list|tuple|Tensor): Shape of the Tensor to be created.
The data type is ``int32`` or ``int64`` . If ``shape`` is a list or tuple,
the elements of it should be integers or Tensors with shape [1].
If ``shape`` is an Variable, it should be an 1-D Tensor .
fill_value(bool|float16|float32|float64|int32|int64|Variable): The constant value
used to initialize the Tensor to be created. If fill_value is an Variable, it must be an 1-D Tensor.
dtype(np.dtype|core.VarDesc.VarType|str, optional): Data type of the output tensor
If ``shape`` is an Tensor, it should be an 1-D Tensor .
fill_value(bool|float|int|Tensor): The constant value
used to initialize the Tensor to be created. If ``fill_value`` is an Tensor, it must be an 1-D Tensor.
dtype(np.dtype|core.VarDesc.VarType|str, optional): Data type of the output Tensor
which can be float16, float32, float64, int32, int64, if dytpe is `None`, the data
type of created tensor is `float32`
type of created Tensor is `float32`
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:
Variable: Tensor which is created according to shape and dtype.
Tensor: Tensor which is created according to ``shape``, ``fill_value`` and ``dtype``.
Raises:
TypeError: The `dtype` must be one of None, bool, float16, float32, float64, int32 and int64.
TypeError: The `shape` must be one of Variable, list and tuple.
TypeError: The ``dtype`` must be one of None, bool, float16, float32, float64, int32 and int64.
TypeError: The ``shape`` must be one of Tensor, list and tuple. The data type of ``shape`` must
be int32 or int64 when the it's a Tensor
Examples:
.. code-block:: python
@ -364,18 +390,18 @@ def full(shape, fill_value, dtype=None, name=None):
#[[0]
# [0]]
# attr shape is a list which contains Variable Tensor.
# attr shape is a list which contains Tensor.
positive_2 = paddle.fill_constant([1], "int32", 2)
data3 = paddle.full(shape=[1, positive_2], dtype='float32', fill_value=1.5)
# [[1.5 1.5]]
# attr shape is an Variable Tensor.
# attr shape is a Tensor.
shape = paddle.fill_constant([2], "int32", 2)
data4 = paddle.full(shape=shape, dtype='bool', fill_value=True)
# [[True True]
# [True True]]
# attr fill_value is an Variable Tensor.
# attr fill_value is a Tensor.
val = paddle.fill_constant([1], "float32", 2.0)
data5 = paddle.full(shape=[2,1], fill_value=val, dtype='float32')
# [[2.0]

14
python/paddle/tensor/manipulation.py
Unescape View File

@ -54,27 +54,27 @@ __all__ = [
def concat(x, axis=0, name=None):
"""
:alias_main: paddle.concat
:alias: paddle.concat,paddle.tensor.concat,paddle.tensor.manipulation.concat
:alias: paddle.tensor.concat, paddle.tensor.manipulation.concat
This OP concatenates the input along the axis.
Args:
x(list): List of input Tensors with data type float16, float32, float64, int32, int64.
All the Tensors in ``x`` must have same data type.
axis(int|Variable, optional): Specify the axis to operate on the input Tensors.
It's a scalar with type ``int`` or a ``Tensor`` with shape [1] and data type ``int32``
or ``int64``. The effective range is [-R, R), where R is Rank(x). When ``axis < 0``,
it works the same way as axis+R. Default is 0.
axis(int|Tensor, optional): Specify the axis to operate on the input Tensors.
It's a scalar with data type int or a Tensor with shape [1] and data type int32
or int64. The effective range is [-R, R), where R is Rank(x). When ``axis < 0``,
it works the same way as ``axis+R``. Default is 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`.
Raises:
TypeError: The dtype of ``x`` must be one of float16, float32, float64, int32 and int64.
TypeError: The ``axis`` must be int or Variable. The dtype of ``axis`` must be int32 or int64 when it's a Tensor.
TypeError: The ``axis`` must be int or Tensor. The dtype of ``axis`` must be int32 or int64 when it's a Tensor.
TypeError: All the Tensors in ``x`` must have the same data type.
Returns:
Variable: A Tensor with the same data type as ``x``.
Tensor: A Tensor with the same data type as ``x``.
Examples:
.. code-block:: python

22
python/paddle/tensor/search.py
Unescape View File

@ -223,27 +223,27 @@ def argmax(input, axis=None, dtype=None, out=None, keepdims=False, name=None):
def index_select(x, index, axis=0, name=None):
"""
:alias_main: paddle.index_select
:alias: paddle.index_select,paddle.tensor.index_select,paddle.tensor.search.index_select
:alias: paddle.tensor.index_select, paddle.tensor.search.index_select
Returns a new tensor which indexes the `input` tensor along dimension `dim` using
the entries in `index` which is a Tensor. The returned tensor has the same number
of dimensions as the original `input` tensor. The dim-th dimension has the same
size as the length of `index`; other dimensions have the same size as in the `input` tensor.
Returns a new tensor which indexes the ``input`` tensor along dimension ``axis`` using
the entries in ``index`` which is a Tensor. The returned tensor has the same number
of dimensions as the original ``x`` tensor. The dim-th dimension has the same
size as the length of ``index``; other dimensions have the same size as in the ``x`` tensor.
Args:
x (Variable): The input tensor variable.The dtype of x can be one of float32, float64, int32, int64.
index (Variable): The 1-D tensor containing the indices to index.the dtype of index can be int32 or int64.
axis (int, optional): The dimension in which we index. Default: if None, the axis is 0.
x (Tensor): The input Tensor to be operated. The data of ``x`` can be one of float32, float64, int32, int64.
index (Tensor): The 1-D Tensor containing the indices to index. The data type of ``index`` must be int32 or int64.
axis (int, optional): The dimension in which we index. Default: if None, the ``axis`` is 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:
Variable: A Tensor with same data type as `input`.
Tensor: A Tensor with same data type as ``x``.
Raises:
TypeError: x must be a Variable and the dtype of x must be one of float32, float64, int32 and int64.
TypeError: index must be a Variable adn the dtype of index must be int32 or int64.
TypeError: ``x`` must be a Tensor and the data type of ``x`` must be one of float32, float64, int32 and int64.
TypeError: ``index`` must be a Tensor and the data type of ``index`` must be int32 or int64.
Examples:
.. code-block:: python

Write Preview
Loading…
Cancel
Save