Index collections

Collections of Index are used throughout ITensors.jl to represent the dimensions of tensors. In general, collections that are recognized and returned by ITensors.jl functions are either Vector of Index or Tuple of Index, depending on the context. For example internally an ITensor has a static number of indices so stores a Tuple of Index, while set operations like commoninds((i, j, k), (j, k, l)) will return a Vector [j, k] since the operation is inherently dynamic, i.e. the number of indices in the intersection can't in general be known before running the code. Vector of Index and Tuple of Index can usually be used interchangeably, but one or the other may be faster depending on the operation being performed.

Priming and tagging

Documentation for priming and tagging collections of Index can be found in the ITensor Priming and tagging section.

Set operations

Documentation for set operations involving Index collections can be found in the ITensor Index collections set operations section.

Subsets

ITensors.getfirstMethod
getfirst(f::Function, is::Indices)

Get the first Index matching the pattern function, return nothing if not found.

source
ITensors.getfirstMethod
getfirst(is::Indices)

Return the first Index in the Indices. If the Indices is empty, return nothing.

source

Iterating

ITensors.eachvalMethod
eachval(is::Index...)
eachval(is::Tuple{Vararg{Index}})

Create an iterator whose values correspond to a Cartesian indexing over the dimensions of the provided Index objects.

source
ITensors.eachindvalMethod
eachindval(is::Index...)
eachindval(is::Tuple{Vararg{Index}})

Create an iterator whose values are Index=>value pairs corresponding to a Cartesian indexing over the dimensions of the provided Index objects.

Example

i = Index(3; tags="i")
j = Index(2; tags="j")
T = random_itensor(j, i)
for iv in eachindval(i, j)
  @show T[iv...]
end
source
ITensors.dirMethod
dir(is::Indices, i::Index)

Return the direction of the Index i in the Indices is.

source