Total spin operator for Electron sites system

[ryanlevy]

To continue Vince’s answer, check out the references here:

MPO construction and compression principles (or: how does OpSum work?)

For construction, I would recommend checking out if you haven’t [0708.1221] Finite automata for caching in matrix product algorithms [0804.2509] Infinite size density matrix renormalization group, revisited And this paper on MPO compression [1909.06341] Local Matrix Product Operators: Canonical Form, Compression, & Control Theory I’ll have to let others comment on the exact compression techniques that are in ITensors

and Miles has nice notes below:

MPO construction and compression principles (or: how does OpSum work?)

This is an interesting topic, for sure, and there has been a lot of theoretical progress on it. The main reference for the system currently used by OpSum is: “Matrix Product Operators, Matrix Product States, and ab initio Density Matrix Renormalization Group algorithms” Chan, Keselman, Nakatani, Li, White https://arxiv.org/abs/1605.02611. For a brief summary of the construction there, I’m attaching the pdf notes below that I wrote for one of the Flatiron postdocs helping us to generalize tha…

---

As a follow up question, would you assume that adding such term to the Hamiltonian for a 2x8 system (max.bond dim. ~2000-4000) would still be a feasible calculation? (I guess I’ll just have to try).

You can always pass MPOs individually (dmrg([H1,H2], psi0;...), and since this is a low “bond dimension” MPO it should not create too much more work compared your “energy” Hamiltonian