Types

template<typename Type>
using row_view = marray_view<Type, 1>

Alias for a 1-dimensional tensor view.

Template Parameters:

Type – The type of the tensor elements.

template<typename Type, typename Allocator = std::allocator<Type>>
using row = marray<Type, 1, Allocator>

Alias for a 1-dimensional tensor.

Template Parameters:
  • Type – The type of the tensor elements.

  • Allocator – An allocator. If not specified, std::allocator<Type> is used.

template<typename Type>
using matrix_view = marray_view<Type, 2>

Alias for a 2-dimensional tensor view.

Template Parameters:

Type – The type of the tensor elements.

template<typename Type, typename Allocator = std::allocator<Type>>
using matrix = marray<Type, 2, Allocator>

Alias for a 2-dimensional tensor.

Template Parameters:
  • Type – The type of the tensor elements.

  • Allocator – An allocator. If not specified, std::allocator<Type> is used.

typedef MARRAY_LEN_TYPE len_type

The integral type used for index lengths.

Typically the same as ptrdiff_t.

typedef MARRAY_STRIDE_TYPE stride_type

The integral type used for index strides.

Typically the same as ptrdiff_t.

struct layout
#include <marray_fwd.hpp>

Type specifying one of the pre-defined tensor layouts.

The pre-defined layouts place tensor elements in continguous locations in memory; more flexible layouts can be achieved by explicitly specifying the strides of the tensor indices.

Subclassed by layout_like< Type, NDim >::no_layout

struct construct
struct index_base
#include <marray_fwd.hpp>

Type specifying one of the pre-defined tensor bases.

The pre-defined bases index all tensor dimensions starting at either zero or one. More flexible bases can be obtained by explicitly specifying the base for each tensor dimension.

Subclassed by base_like< Type, NDim >::no_base

struct construct