Remove EOM- from prop name when irrelevant #2533
Conversation
Computing properly and not giving the expected |
| @@ -96,11 +96,11 @@ for transition, data in test_data.items(): # TEST | |||
| for property, value in data.items(): # TEST | |||
| # Change these numbers to 6 ASAP # TEST | |||
| if "EINSTEIN" not in property: # TEST | |||
| compare_values(value, wfn.variable(f"CC {transition} {property}"), 6, f"CC {transition} {property}") # TEST | |||
| compare_values(value, wfn.variable(f"CCSD {transition} {property}"), 6, f"CCSD {transition} {property}") # TEST | |||
| compare_values(value, wfn.variable(f"CC {transition} {property}"), 4, f"CC {transition} {property}") # TEST | |||
There was a problem hiding this comment.
any clue what the above # Change these numbers to 6 ASAP is about, here? it's suspect if these oscillator strengths aren't matching to at least 1e-6.
There was a problem hiding this comment.
It's unjustifiable to check these quantities to 1e-6 when cc refuses to enforce lambda convergence any tighter than 1e-4. I have plans to fix this. See cclambda's sin list.
But that scale of refactoring is outside the scope of this review.
There was a problem hiding this comment.
wow. that is highly irregular, and definitely a problem. I'll be looking forward to these refactors. thanks!
* Remove EOM- from prop name when irrelevant * Loosen cc55
* Remove EOM- from prop name when irrelevant * Loosen cc55
Description
PR #2022 labeled ground-state properties with the method name... even when the method name included "EOM-". So even though
ccsdandeom-ccsdcompute the same ground state dipole, they weren't given the same name. This PR fixes that.More fixes to come, but the next PR will be heftier.
Obligatory @bgpeyton ping.
Checklist
eomtests passStatus