Lori A. Burns
Lori A. Burns
Thanks very much! (1b) Still on the Datum part, now it gets through all the serializing all the real ndarrays, but it balks on the complex (freq for transition states)....
1b definitely resolved, thanks! (1d) What's your view on numpy floats? I can easily solve this in my code by casting to float, but as-is, one can create a Datum...
Confirmed (1d) fixed! thanks, @Lnaden
Nice strategy with the deprecation warning on `.dict` and `.json`, btw. The first time through the full psi4 test suite, I triggered 89k of them :-) . Now fixed except...
> Thanks. I tried to preserve the API on our side as much as possible while keeping the deprecation warning. I can remove the warning on our side and just...
> Is this effort still active? ... It's still active, and it works (except for one line with optking). And downstream v2-proper PRs are largely ready for QCEngine and Psi4....
Thanks for the update, @mattwthompson . After I get through these next papers I'll try a dual v1/v2 version of qcel. But I bet one wants a triple version for...
I'll make sure to bring this up to @bennybp to see if the database is ready for a type change.
actually will be fixed by #340
Ah yes, in the beginning it was requested that fragments be allowed to be noncontiguous. So the good news is that I wrote the hard part (molparse validator) to accommodate....