Now, I am coding 1-site DMRG with subspace expansion method (or 3S method) in Julia, which includes the additional step of adding perturbation term by increasing virtual bond dimension. Initially, I encountered an issue during testing while using the combiner function to merge Indexes considering qu…

Could you share some minimal runnable example code demonstrating the issue you are seeing?

There is a directsum function available in ITensors.jl for direct summing two (or more) ITensors.

Sorry for late due to lunch Test code is as follows. N=10 sites = siteinds(“S=1”, N; conserve_qns= true ) os = OpSum() for j=1:N-1 os += “Sz”,j,“Sz”,j+1 os += 1/2,“S+”,j,“S-”,j+1 os += 1/2,“S-”,j,“S+”,j+1 end H = MPO(os, sites) states = [isodd(n) ? “Up” : “Dn” for n in 1:N] χ = [ i for i…

When i command that directsum(A, B), it gives error: MethodError: no method matching directsum(::ITensor, ::ITensor) (where A, B are 3-legs tensors that two indices are shared and a index are different)

Please look at the documentation for directsum by typing ?directum at the Julia prompt for instructions on how to use it, you have to specify which indices should get direct summed.

Q1. and Q2. That index is constructed as part of the OpSum to MPO conversion code. The code creates indices with that structure to keep track of operators in different QN sectors. It isn’t a problem that the sectors are split in that way, in fact it can lead to MPO tensors that are more sparse which…

Here is the running code using ITensors N=10; sites = siteinds("S=1", N; conserve_qns= true ); os = OpSum(); for j=1:N-1 os += "Sz",j,"Sz",j+1 os += 1/2,"S+",j,"S-",j+1 os += 1/2,"S-",j,"S+",j+1 end H = MPO(os, sites); states = [isodd(n) ? "Up" : "Dn" for n in 1:N]; χ = [ i for i in…

I appreciate your clear explanations which are easy to understand. Your answers to my beginner-level questions have been very helpful. Thank you for addressing my inquiries so well. ITensors seems to be a very impressive code.

There is an improved version of the 3S algorithm which is much faster and is actually easier to implement than the old 3S algorithm. For a new code, I would recommend implementing the new scheme (“post-expansion”) instead of 3S, as it has many advantages. It is easier to use because you don’t need…

Combiner built-in function issue for virtual bond index from MPO with quantum number conservation

ITensor Julia Questions

mtfishman April 2, 2024, 11:31am 8

Please format your code, see the post Please Read: Make It Easier to Help You.

Topic		Replies	Views
Sum of two MPOs ITensor Julia Questions	8	653	March 24, 2023
Is some compression on sum of MPOs done automatically? ITensor Julia Questions	2	157	September 18, 2024
Making ITensor merge arbitary tensors with named index values ITensor Julia Questions	1	52	October 31, 2025
Fixed bond dimension operations ITensor Julia Questions julia , mpo	9	151	January 7, 2025
Inconsistent behaviour of combiner with and without qns ITensor Julia Questions	0	44	August 13, 2025

Combiner built-in function issue for virtual bond index from MPO with quantum number conservation

Related topics