search

ACEsuit / ACEpsi.jl

ACE wave function parameterizations
MIT License
2 stars 0 forks source link

WIP: Envelope #12

Closed CheukHinHoJerry closed 1 year ago

CheukHinHoJerry commented 1 year ago

Just finish up the gradp for decay and lap of decay with new tests included. Not sure if the testing result is correct since the gradp is linear w.r.t parameter of the decay because of the ln(exp(...)) (if I understand correctly).

I also implements the set/get_param for enveolpe and BFwf with testing. Test for compatibility with ACESchrodinger is also cleaned up so that we don't have to compare it with both of the code running.

@cortner Please review if you have time.

cortner commented 1 year ago

i notice this does not follow the convention for parameters and for gradients that we use in ACESchrodinger. That's not necessarily a problem if you write some interface/wrappers. But eventually we need to do this here, and do it consistently.

cortner commented 1 year ago

If you are ok with that, then please file an issue so we remember this for the future and then I'm happy for you to merge this PR. We can next discuss how to integrate this new code into ACESchrodinger.

CheukHinHoJerry commented 1 year ago

Thanks a lot for reviewing. I will merge the code tmr to the main branch and close this PR after merching.