GitHub-Proxy/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2025-07-07 05:31:48 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix docs build.

Browse Source
This commit is contained in:
Antonio Sánchez 2025-07-02 22:10:33 +00:00
parent f169c13d8e
commit cc0be00435
1 changed files with 18 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
unsupported/Eigen/CXX11/src/Tensor/README.md
View File

@ -92,7 +92,7 @@ TensorMap<Tensor<float, 1>> t_12(t_4x3.data(), 12);
#### Class TensorRef #### Class TensorRef
See [Assigning to a TensorRef.](#assigning-to-a-tensorref) See **Assigning to a `TensorRef`**.
## Accessing Tensor Elements ## Accessing Tensor Elements
@ -602,8 +602,8 @@ std::cout << "Size: " << a.size();
A few operations provide `dimensions()` directly, A few operations provide `dimensions()` directly,
e.g. `TensorReslicingOp`. Most operations defer calculating dimensions e.g. `TensorReslicingOp`. Most operations defer calculating dimensions
until the operation is being evaluated. If you need access to the dimensions until the operation is being evaluated. If you need access to the dimensions
of a deferred operation, you can wrap it in a TensorRef (see of a deferred operation, you can wrap it in a `TensorRef` (see
[Assigning to a TensorRef.](#assigning-to-a-tensorref)), which provides **Assigning to a TensorRef** above), which provides
`dimensions()` and `dimension()` as above. `dimensions()` and `dimension()` as above.
`TensorRef` can also wrap the plain `Tensor` types, so this is a useful idiom in `TensorRef` can also wrap the plain `Tensor` types, so this is a useful idiom in
@ -635,7 +635,7 @@ is not initialized.
Eigen::TensorFixedSize<float, Sizes<3, 4>> a; Eigen::TensorFixedSize<float, Sizes<3, 4>> a;
std::cout << "Rank: " << a.rank() << endl; std::cout << "Rank: " << a.rank() << endl;
// Rank: 2 // Rank: 2
std::cout << "NumRows: " << a.dimension(0) std::cout << "NumRows: " << a.dimension(0)
<< " NumCols: " << a.dimension(1) << endl; << " NumCols: " << a.dimension(1) << endl;
// NumRows: 3 NumCols: 4 // NumRows: 3 NumCols: 4
``` ```
@ -848,7 +848,7 @@ These can be chained: you can apply another `Tensor` Operation to the value
returned by the method. returned by the method.
The chain of Operation is evaluated lazily, typically when it is assigned to a The chain of Operation is evaluated lazily, typically when it is assigned to a
tensor. See [Controlling When Expression are Evaluated](#controlling-when-expression-are-evaluated) for more details about tensor. See **Controlling When Expression are Evaluated** for more details about
their evaluation. their evaluation.
### (Operation) constant(const Scalar& val) ### (Operation) constant(const Scalar& val)
@ -858,7 +858,7 @@ where all elements have the value `val`.
This is useful, for example, when you want to add or subtract a constant from a This is useful, for example, when you want to add or subtract a constant from a
tensor, or multiply every element of a tensor by a scalar. tensor, or multiply every element of a tensor by a scalar.
However, such operations can also be performed using operator overloads (see [operator+](#operation-operator-scalar-s)). However, such operations can also be performed using operator overloads (see `operator+`).
```cpp ```cpp
@ -927,7 +927,7 @@ std::cout << "b\n" << b << "\n\n";
// a // a
// 1 1 1 // 1 1 1
// 1 1 1 // 1 1 1
// //
// b // b
// -1 -1 -1 // -1 -1 -1
// -1 -1 -1 // -1 -1 -1
@ -1010,7 +1010,7 @@ std::cout << "b" << endl << b << endl << endl;
// a // a
// 0 1 8 // 0 1 8
// 27 64 125 // 27 64 125
// //
// b // b
// 0 1 2 // 0 1 2
// 3 4 5 // 3 4 5
@ -1031,7 +1031,7 @@ std::cout << "scaled_a\n" << scaled_a << "\n";
// a // a
// 1 2 3 // 1 2 3
// 4 5 6 // 4 5 6
// //
// scaled_a // scaled_a
// 2 4 6 // 2 4 6
// 8 10 12 // 8 10 12
@ -1048,8 +1048,8 @@ Divides every element in the tensor by `s`.
### (Operation) operator% (Scalar s) ### (Operation) operator% (Scalar s)
Computes the element-wise modulus (remainder) of each tensor element divided by `s` Computes the element-wise modulus (remainder) of each tensor element divided by `s`
**Only integer types are supported.** **Only integer types are supported.**
For floating-point tensors, implement a [unaryExpr](#operation-unaryexprcustomunaryop-func) using `std::fmod`. For floating-point tensors, implement a `unaryExpr` using `std::fmod`.
### (Operation) cwiseMax(Scalar threshold) ### (Operation) cwiseMax(Scalar threshold)
Returns the coefficient-wise maximum between two tensors. Returns the coefficient-wise maximum between two tensors.
@ -1203,10 +1203,10 @@ The following boolean operators are supported:
* `operator>=(const OtherDerived& other)` * `operator>=(const OtherDerived& other)`
* `operator==(const OtherDerived& other)` * `operator==(const OtherDerived& other)`
* `operator!=(const OtherDerived& other)` * `operator!=(const OtherDerived& other)`
as well as bitwise operators: as well as bitwise operators:
* `operator&(const OtherDerived& other)` * `operator&(const OtherDerived& other)`
* `operator|(const OtherDerived& other)` * `operator|(const OtherDerived& other)`
* `operator^(const OtherDerived& other)` * `operator^(const OtherDerived& other)`
@ -1448,7 +1448,7 @@ std::cout << "Flat argmax index: " << argmax_flat();
### (Operation) argmin(const Dimensions& reduction_dim) ### (Operation) argmin(const Dimensions& reduction_dim)
### (Operation) argmin() ### (Operation) argmin()
See [argmax](#operation-argmaxconst-dimensions-reduction_dim) See `argmax`.
### (Operation) reduce(const Dimensions& reduction_dims, const Reducer& reducer) ### (Operation) reduce(const Dimensions& reduction_dims, const Reducer& reducer)
@ -1953,7 +1953,7 @@ std::cout << "b\n" << b << "\n";
### (Operation) roll(const Rolls& shifts) ### (Operation) roll(const Rolls& shifts)
Returns a tensor with the elements **circularly shifted** (like bit rotation) along one or more dimensions. Returns a tensor with the elements **circularly shifted** (like bit rotation) along one or more dimensions.
For each dimension `i`, the content is shifted by `shifts[i]` positions: For each dimension `i`, the content is shifted by `shifts[i]` positions:
@ -2277,7 +2277,7 @@ std::cout << "b\n" << b << "\n";
``` ```
### (Operation) eval() ### (Operation) eval()
See [Calling eval()](#calling-eval) See **Calling eval()**.
@ -2340,7 +2340,7 @@ For example `Tensor<T, N>::maximum()` returns a `Tensor<T, 0>`.
Similarly, the inner product of 2 1d tensors (through contractions) returns a 0d tensor. Similarly, the inner product of 2 1d tensors (through contractions) returns a 0d tensor.
The scalar value can be extracted as explained in [Reduction along all dimensions](#reduction-along-all-dimensions). The scalar value can be extracted as explained in **Reduction along all dimensions**.
## Limitations ## Limitations
@ -2349,4 +2349,4 @@ The scalar value can be extracted as explained in [Reduction along all dimension
compiler that supports cxx11. It is limited to only 5 for older compilers. compiler that supports cxx11. It is limited to only 5 for older compilers.
* The `IndexList` class requires a cxx11 compliant compiler. You can use an * The `IndexList` class requires a cxx11 compliant compiler. You can use an
array of indices instead if you don't have access to a modern compiler. array of indices instead if you don't have access to a modern compiler.
* On GPUs only floating point values are properly tested and optimized for. * On GPUs only floating point values are properly tested and optimized for.