search

ICAMS / calphy

A Python library and command line interface for automated free energy calculations
Other
68 stars 21 forks source link

Update find_w call sites #114

Closed pmrv closed 7 months ago

pmrv commented 7 months ago

ca61abc changed its signature but did not change how it is called in liquid and alchemical modes.

pmrv commented 7 months ago

So sadly even with this I get all nan's for the liquid free energy. I've traced it until integrate_rs. If I call that function manually it gives me back finite numbers, but in the actual run (see attachment) it's all nan. A hint is that in the report.yaml the free energy (of the NVT reference, I guess) is already nan. But I'm stuck here because I don't know where that reference energy comes from. Can it be some problem with uhlbeck-ford model? I've check the traj.melt snapshot, but it seems ok.

liquid.gz

srmnitc commented 7 months ago

Thanks @pmrv. I merged this in. The issue in liquid was coming from integrator trying to take log(0), fixed in #117. This happened when the pair style had multiple elements, but only one was present in the structure. This would work now. I have also released v 1.3.6 on conda and pip.

srmnitc commented 7 months ago

Just FYI: I found this from error.out, where it was complaining about division by zero.