Repository navigation
PySCFTheory: use total density for the point-charge gradient of unrestricted SCF - #527
Merged
RagnarB83 merged 1 commit intoOct 20, 2026
Conversation
…tricted SCF For UHF/UKS, make_rdm1() returns (dm_alpha, dm_beta) and the "new" point-charge gradient code kept only dm[0]. grad_hcore_mm is linear in the density, so the electronic contribution to the forces on the MM charges was missing all beta electrons, while the energy and the QM-atom gradient were correct. Use dm[0] + dm[1]. Adds a finite-difference test of the point-charge gradient for RHF and UHF. Co-Authored-By: Claude Opus 5.5 <[email protected]>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #526.
For UHF/UKS,
make_rdm1()returns(dm_alpha, dm_beta), and the point-charge gradient (PC_gradient_code = "new") kept onlydm[0].grad_hcore_mmis linear in the density, so the electronic contribution to the forces on the MM charges was missing every beta electron. This PR usesdm[0] + dm[1]. The energy and the QM-atom gradient were already correct and are unchanged, and closed-shell runs (dm.ndim == 2) are unaffected.Changes
ash/interfaces/interface_pyscf.py: one line (+ comment).ash/tests/test_pyscf_pointcharge_gradient.py(new): comparespcgradwith central finite differences of the energy for RHF (H2O) and UHF (OH radical) next to three point charges. It needs only CPU PySCF (skipped if PySCF is missing) and takes about 3 s.Testing
ash/tests/test_pyscf_scf_convergence.pystill passes.🤖 Generated with Claude Code