Repository navigation
CH2I2/CH2I data for stdsuite - #526
Merged
Merged
Conversation
Adds non-equilibrium C2v CH2I2 and CH2I to std_molecules (I with 28-e def2
ECP), def2-svp generics, and CONV-{AE,FC}-CONV, CONV-{AE,FC}-DF, and
DF-{AE,FC}-DF blocks (rhf/uhf sd; rohf sd+sc) with HF, MP2, CCSD, and
CCSD(T) reference values. All values from psi4; every analytic gradient
stored was checked against findif-5 of energies (Cartesian steps) to
<=1.5e-7, analytic HF hessians against findif, and cross-module values
(occ/dfmp2/dfocc/ccenergy/fnocc) against each other. Provenance noted inline.
No values for existing systems changed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Libecpint had four accuracy defects in type-2 integrals for basis shells off the ECP centre (unrolled-kernel literals truncated to 6 digits, a wrong closed form for radial (1,1,10), an under-sized fallback quadrature window for high orders, closed-form cancellation for diffuse primitives). All 162 ECP reference values are regenerated with Psi4 linked to the fixed Libecpint, with the same generators, settings and provenance notes as before; nothing outside the ECP blocks changes. Changes vs the previous values: energies <=4e-8, gradients <=9e-8, analytic HF Hessians 1.2e-6 (ch2i2) / 6.2e-7 (ch2i), MP2 findif Hessians <=2.8e-6; findif-of-gradient DF/ROHF HF Hessians move up to 3.6e-5 (run-to-run findif noise for I). Every stored analytic gradient agrees with Cartesian-step findif of energies (<=2.8e-7 for DF, <=6e-8 otherwise) and every analytic Hessian with findif of analytic gradients (<=3.6e-6). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Previous refs were findif-5 of DF analytic gradients at disp_size 0.005, which carry up to ~5e-5 noise for iodine (ill-conditioned jkfit metric plus tiny mass-weighted steps on I). Analytic DF hessians, newly reachable with a libint2 having 3eri2/2eri2 AM >= 5, agree with findif at disp_size 0.04 to 5e-6 (rhf) and 3e-7 (uhf). ROHF stays findif. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Codecov Report✅ All modified and coverable lines are covered by tests. 🚀 New features to boost your workflow:
|
bennybp
approved these changes
Oct 16, 2026
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
2 participants
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.
Description
Changelog description
Status