Skip to content

Improve Orca Optimization and Frequency Convergence Settings#888

Open
Ophir-Weisz wants to merge 2 commits into
mainfrom
orca_opt_freq
Open

Improve Orca Optimization and Frequency Convergence Settings#888
Ophir-Weisz wants to merge 2 commits into
mainfrom
orca_opt_freq

Conversation

@Ophir-Weisz
Copy link
Copy Markdown

Changes

  • Standardized optimization convergence to TightOpt for all opt jobs (was NormalOpt)
  • Enhanced frequency calculations to use defgrid3 instead of defgrid2 (consistent with fine opt)
  • Improved TS optimization with Recalc_Hess 5 in %geom block for periodic Hessian recalculation
  • Added comprehensive unit tests to validate all changes

Benefits

  • Better convergence reliability across different molecular systems
  • Consistent DFT grid quality between optimization and frequency steps
  • Improved transition state geometry optimization

Copy link
Copy Markdown

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR updates ARC’s ORCA adapter defaults to improve optimization convergence behavior, align DFT grid settings between fine optimization and frequency calculations, and enhance TS optimizations with periodic Hessian recalculation. It also adds unit tests to validate the new ORCA input generation behavior.

Changes:

  • Set ORCA default optimization convergence to TightOpt (previously NormalOpt).
  • Default DFT grid selection to defgrid3 for freq/optfreq (and for fine optimizations), while keeping regular opt at defgrid2.
  • For TS optimizations (OptTS), add Recalc_Hess 5 to the %geom block, and add corresponding unit tests for grid/convergence/TS Hessian settings.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File Description
arc/settings/settings.py Updates ORCA default optimization convergence (opt_convergence) to TightOpt.
arc/job/adapters/orca.py Adjusts default DFT grid behavior for freq/optfreq and adds Recalc_Hess 5 for OptTS.
arc/job/adapters/orca_test.py Adds unit tests covering DFT grid defaults, TightOpt presence, and TS %geom Hessian recalc settings.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread arc/job/adapters/orca.py
if self.level.method_type == 'dft':
input_dict['method_class'] = 'KS'
# DFT grid must be the same for both opt and freq.
# DFT grid must be the same for both fine_opt and freq.
Comment thread arc/job/adapters/orca.py
self.add_to_args(val="""
%geom
Calc_Hess true # calculation of the exact Hessian before the first opt step
Recalc_Hess 5
@codecov
Copy link
Copy Markdown

codecov Bot commented May 13, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 60.50%. Comparing base (5f2138a) to head (77a5248).
⚠️ Report is 6 commits behind head on main.

Additional details and impacted files
@@            Coverage Diff             @@
##             main     #888      +/-   ##
==========================================
+ Coverage   60.49%   60.50%   +0.01%     
==========================================
  Files         103      103              
  Lines       31153    31165      +12     
  Branches     8120     8126       +6     
==========================================
+ Hits        18845    18857      +12     
+ Misses       9966     9948      -18     
- Partials     2342     2360      +18     
Flag Coverage Δ
functionaltests 60.50% <ø> (+0.01%) ⬆️
unittests 60.50% <ø> (+0.01%) ⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants