TensorFlow教程之API DOC 6.1.6 Class tensorflow::Tensor-阿里云开发者社区

开发者社区> 人工智能> 正文

TensorFlow教程之API DOC 6.1.6 Class tensorflow::Tensor

简介:

本文档为TensorFlow参考文档,本转载已得到TensorFlow中文社区授权。


Class tensorflow::Tensor

Represents an n-dimensional array of values.

Member Summary

Member Details

tensorflow::Tensor::Tensor()

Default Tensor constructor. Creates a 1-dimension, 0-element float tensor.

tensorflow::Tensor::Tensor(DataType type, const TensorShape &shape)

Creates a Tensor of the given type and shape.

The underlying buffer is allocated using a CPUAllocator.

tensorflow::Tensor::Tensor(Allocator *a, DataType type, const TensorShape &shape)

Creates a tensor with the input type and shape, using the allocator a to allocate the underlying buffer.

a must outlive the lifetime of this Tensor .

tensorflow::Tensor::Tensor(DataType type)

Creates an uninitialized Tensor of the given data type.

tensorflow::Tensor::Tensor(const Tensor &other)

tensorflow::Tensor::~Tensor()

Copy constructor.

DataType tensorflow::Tensor::dtype() const

Returns the data type.

const TensorShape& tensorflow::Tensor::shape() const

Returns the shape of the tensor.

int tensorflow::Tensor::dims() const

Convenience accessor for the tensor shape.

For all shape accessors, see comments for relevant methods of TensorShape in tensor_shape.h.

int64 tensorflow::Tensor::dim_size(int d) const

Convenience accessor for the tensor shape.

int64 tensorflow::Tensor::NumElements() const

Convenience accessor for the tensor shape.

bool tensorflow::Tensor::IsSameSize(const Tensor &b) const

bool tensorflow::Tensor::IsInitialized() const

Has this Tensor been initialized?

size_t tensorflow::Tensor::TotalBytes() const

Returns the estimated memory usage of this tensor.

Tensor& tensorflow::Tensor::operator=(const Tensor &other)

Assign operator. This tensor shares other's underlying storage.

bool tensorflow::Tensor::CopyFrom(const Tensor &other, const TensorShape &shape) TF_MUST_USE_RESULT

Copy the other tensor into this tensor and reshape it.

This tensor shares other's underlying storage. Returns true iff other.shape() has the same number of elements of the given shape.

Tensor tensorflow::Tensor::Slice(int64 dim0_start, int64 dim0_limit) const

Slice this tensor along the 1st dimension.

I.e., the returned tensor satisifies returned[i, ...] == this[dim0_start + i, ...]. The returned tensor shares the underlying tensor buffer with this tensor.

NOTE: The returned tensor may not satisfies the same alignment requirement as this tensor depending on the shape. The caller must check the returned tensor's alignment before calling certain methods that have alignment requirement (e.g., flat()tensor()).

REQUIRES: dims() >= 1 REQUIRES: 0 <= dim0_start <= dim0_limit <= dim_size(0)

bool tensorflow::Tensor::FromProto(const TensorProto &other) TF_MUST_USE_RESULT

Parse other and construct the tensor.

Returns true iff the parsing succeeds. If the parsing fails, the state of *this is unchanged.

bool tensorflow::Tensor::FromProto(Allocator *a, const TensorProto &other) TF_MUST_USE_RESULT

void tensorflow::Tensor::AsProtoField(TensorProto *proto) const

Fills in proto with *this tensor's content.

AsProtoField() fills in the repeated field for proto.dtype(), while AsProtoTensorContent() encodes the content in proto.tensor_content() in a compact form.

void tensorflow::Tensor::AsProtoTensorContent(TensorProto *proto) const

TTypes<T>::Vec tensorflow::Tensor::vec()

Return the tensor data as an Eigen::Tensor with the type and sizes of this Tensor.

Use these methods when you know the data type and the number of dimensions of the Tensor and you want an Eigen::Tensor automatically sized to the Tensor sizes. The implementation check fails if either type or sizes mismatch.

Example:

```c++ typedef float T; Tensor my_mat(...built with Shape{rows: 3, cols: 5}...); auto mat = my_mat.matrix(); // 2D Eigen::Tensor, 3 x 5. auto mat = my_mat.tensor(); // 2D Eigen::Tensor, 3 x 5. auto vec = my_mat.vec(); // CHECK fails as my_mat is 2D. auto vec = my_mat.tensor(); // CHECK fails as my_mat is 2D. auto mat = my_mat.matrix();// CHECK fails as type mismatch.,>,>


#### `TTypes<T>::Matrix tensorflow::Tensor::matrix()` <a class="md-anchor" id="TTypes_T_Matrix_tensorflow_Tensor_matrix"></a>





#### `TTypes< T, NDIMS >::Tensor tensorflow::Tensor::tensor()` <a class="md-anchor" id="TTypes_T_NDIMS_Tensor_tensorflow_Tensor_tensor"></a>





#### `TTypes<T>::Flat tensorflow::Tensor::flat()` <a class="md-anchor" id="TTypes_T_Flat_tensorflow_Tensor_flat"></a>

Return the tensor data as an `Eigen::Tensor` of the data type and a specified shape.

These methods allow you to access the data with the dimensions and sizes of your choice. You do not need to know the number of dimensions of the Tensor to call them. However, they `CHECK` that the type matches and the dimensions requested creates an `Eigen::Tensor` with the same number of elements as the tensor.

Example:

```c++ typedef float T;
Tensor my_ten(...built with Shape{planes: 4, rows: 3, cols: 5}...);
// 1D Eigen::Tensor, size 60:
auto flat = my_ten.flat<T>();
// 2D Eigen::Tensor 12 x 5:
auto inner = my_ten.flat_inner_dims<T>();
// 2D Eigen::Tensor 4 x 15:
auto outer = my_ten.shaped<T, 2>({4, 15});
// CHECK fails, bad num elements:
auto outer = my_ten.shaped<T, 2>({4, 8});
// 3D Eigen::Tensor 6 x 5 x 2:
auto weird = my_ten.shaped<T, 3>({6, 5, 2});
// CHECK fails, type mismatch:
auto bad   = my_ten.flat<int32>();

TTypes<T>::UnalignedFlat tensorflow::Tensor::unaligned_flat()

TTypes<T>::Matrix tensorflow::Tensor::flat_inner_dims()

Returns the data as an Eigen::Tensor with 2 dimensions, collapsing all Tensor dimensions but the last one into the first dimension of the result.

TTypes<T>::Matrix tensorflow::Tensor::flat_outer_dims()

Returns the data as an Eigen::Tensor with 2 dimensions, collapsing all Tensor dimensions but the first one into the last dimension of the result.

TTypes< T, NDIMS >::Tensor tensorflow::Tensor::shaped(gtl::ArraySlice< int64 > new_sizes)

TTypes< T, NDIMS >::UnalignedTensor tensorflow::Tensor::unaligned_shaped(gtl::ArraySlice< int64 > new_sizes)

TTypes< T >::Scalar tensorflow::Tensor::scalar()

Return the Tensor data as a TensorMap of fixed size 1: TensorMap<TensorFixedSize<T, 1>>.

Using scalar() allows the compiler to perform optimizations as the size of the tensor is known at compile time.

TTypes<T>::ConstVec tensorflow::Tensor::vec() const

Const versions of all the methods above.

TTypes<T>::ConstMatrix tensorflow::Tensor::matrix() const

TTypes< T, NDIMS >::ConstTensor tensorflow::Tensor::tensor() const

TTypes<T>::ConstFlat tensorflow::Tensor::flat() const

TTypes<T>::UnalignedConstFlat tensorflow::Tensor::unaligned_flat() const

TTypes<T>::ConstMatrix tensorflow::Tensor::flat_inner_dims() const

TTypes<T>::ConstMatrix tensorflow::Tensor::flat_outer_dims() const

TTypes< T, NDIMS >::ConstTensor tensorflow::Tensor::shaped(gtl::ArraySlice< int64 > new_sizes) const

TTypes< T, NDIMS >::UnalignedConstTensor tensorflow::Tensor::unaligned_shaped(gtl::ArraySlice< int64 > new_sizes) const

TTypes< T >::ConstScalar tensorflow::Tensor::scalar() const

string tensorflow::Tensor::SummarizeValue(int64 max_entries) const

Render the first max_entries values in *this into a string.

string tensorflow::Tensor::DebugString() const

A human-readable summary of the tensor suitable for debugging.

void tensorflow::Tensor::FillDescription(TensorDescription *description) const

Fill in the TensorDescription proto with metadata about the tensor that is useful for monitoring and debugging.

StringPiece tensorflow::Tensor::tensor_data() const

Returns a StringPiece mapping the current tensor's buffer.

The returned StringPiece may point to memory location on devices that the CPU cannot address directly.

NOTE: The underlying tensor buffer is refcounted, so the lifetime of the contents mapped by the StringPiece matches the lifetime of the buffer; callers should arrange to make sure the buffer does not get destroyed while the StringPiece is still used.

REQUIRES: DataTypeCanUseMemcpy( dtype() ).

版权声明:本文内容由阿里云实名注册用户自发贡献,版权归原作者所有,阿里云开发者社区不拥有其著作权,亦不承担相应法律责任。具体规则请查看《阿里云开发者社区用户服务协议》和《阿里云开发者社区知识产权保护指引》。如果您发现本社区中有涉嫌抄袭的内容,填写侵权投诉表单进行举报,一经查实,本社区将立刻删除涉嫌侵权内容。

分享:
人工智能
使用钉钉扫一扫加入圈子
+ 订阅

了解行业+人工智能最先进的技术和实践,参与行业+人工智能实践项目

其他文章