Return none spin summed rdms

[ncrubin]

RDMs returned from pyscf are spin summed. Use https://sunqm.github.io/pyscf/_modules/pyscf/fci/direct_spin1.html#FCISolver.make_rdm12s

to return non-spin summed versions. Recombine spin-separated RDMs like run_psi4 to get spin-orbital rdms before returning to OpenFermion.

[anishbhujbal7]

Hi @ncrubin,

I’m interested in working on this issue if it is still available.
I came across the issue while looking for a contribution related to OpenFermion and quantum chemistry. I noticed that the PySCF direct_spin1.py link referenced in the original issue is no longer available at that URL, so I wanted to check whether the proposed approach is still the preferred way to obtain the non-spin-summed RDMs with the current PySCF API.

If the issue is still relevant, I’d be happy to investigate the current PySCF implementation, update the approach as needed, and work on the corresponding tests.

Would it be possible to assign this issue to me if no one is currently working on it?

Thanks!

[ncrubin]

Hi, Sure! I see the link is broken. The new link is probably here (https://github.com/pyscf/pyscf/blob/master/pyscf/fci/direct_spin1.py) though it has been awhile since I looked. The goal was to get spinorbital RDMs. You can probably also use the example

https://github.com/pyscf/pyscf/blob/master/examples/fci/31-apply_2nd_quantized_op.py

but it will be much slower

[anishbhujbal7]

Hi @ncrubin,

I've finished implementing this using direct_spin1.make_rdm12s to get the non-spin-summed RDMs.
I also fixed the slicing in test_accessing_rdm() to properly account for the cross-spin blocks ($\alpha\beta\beta\alpha$ and $\beta\alpha\alpha\beta$), and all tests are passing locally now.
Should I go ahead and open a PR for review?
Thanks!

[ncrubin]

sure!