mtfishman/tensors.jl

Testing out a design for ITensor using an internal Tensor type.

Tensors.jl

Testing out a design for ITensor using an internal Tensor type.

The Tensor class is meant to have a Julia AbstractArray interface, but with various types of storage and indices (i.e. the storage could be Diag or BlockSparse and the indices could be tuples of integers or ITensor Indices).

ITensor functions are written by creating Tensor objects from their storage and indices and then calling Tensor functions, using a function-barrier approach.

Here is a sample of some current capabilities:

include("itensor.jl")

#
# Test the Tensor class
#

TA = Tensor(Dense([1 2; 3 4]),(2,2))
TB = Tensor(Diag([1,2]),(2,2))

@show TA[1,2] == 2
@show TB[2,2] == 2
@show TB[1,2] == 0
@show permutedims(TA,(2,1))
@show TC = TA+TB
@show add_permute(TA,TB,(2,1))

#
# Test the ITensor class
#

i = Index(2,"i")
j = Index(2,"j")

A = ITensor([1 2; 3 4],i,j)
B = diagITensor([1,2],i,j)

@show A[i(1),j(2)]
@show A[j(2),i(1)]
@show B[i(1),j(1)]
@show B[i(1),j(2)]

@show Ap = permute(A,j,i)

@show A[i(1),j(2)] == Ap[i(1),j(2)]

@show C = A+Ap

This code is in the run.jl script in the repository.