Describe the bug
For b97-d3bj and wb97m-d3bj methods, an energy calculation seems to be divided into b97, wb97m plus the d3bj dispersion correction, and this results in psi4 throwing an input_error saying method doesn't exist since there are no functionals b97 and wb97m separately. Not sure where the split happens since qcengine works fine with the above functionals.
To Reproduce
Not sure how to reproduce this. On this single point energies dataset, all runs under those two specs fail. Here is one error message along with the input it takes, where we can see in the qcspec under model the method is wb97m, which is invalid
Describe the bug For
b97-d3bj
andwb97m-d3bj
methods, an energy calculation seems to be divided intob97
,wb97m
plus thed3bj
dispersion correction, and this results in psi4 throwing an input_error saying method doesn't exist since there are no functionalsb97
andwb97m
separately. Not sure where the split happens since qcengine works fine with the above functionals.To Reproduce Not sure how to reproduce this. On this single point energies dataset, all runs under those two specs fail. Here is one error message along with the input it takes, where we can see in the qcspec under model the method is
wb97m
, which is invalidCode to retrieve the error
Output of this code snippet
Expected behavior
A local fractal run with one of the above specs works fine so it is a bit confusing what's the source of error is