Closed mattiafelice-palermo closed 6 months ago
Hi Mattia,
Thanks a lot for bringing this up. As you suggested it must have been a recent change in one of the dependencies - sign that I should be adding more tests! I have now updated the code to fix this issue. Hopefully the rest of your experience with qforce will be smoother, feel free to ask if you have any issues/questions!
Cheers, Selim
Perfect thank you very much for your help, it's much appreciated!
Mattia
Dear Selim,
first of all thank you very much for sharing qforce with the opensource community.
I'm trying to reproduce the propane example before embarking on more complex molecules, but I cannot move past the QM dihedral calculations.
After performing, the dihedral QM PES scan, I run
qforce propane
, but unfortunately the program halts with the following traceback:I did some debugging, and found the issue to be in the
subset
class method from themolecule/non_bonded.py
module. At line 90, there's an if clause to see iffragment charges
has value (if frag_charges != []:
), but it seems the code attempts to compare a numpy array with a python list, resulting in a ValueError.I modified the code converting on the fly the
frag_charges
numpy array into a list just for the if clause, and with this modification qforce runs as expected and finishes the job gracefully.It seems strange that this kind of bug is out in the open, perhaps there has been recent numpy library changed that became more restrictive and prevent the type mixing as shown above? Or maybe I'm just doing something wrong...
Thank you and I hope this is useful.
Mattia