Open retinavr opened 3 years ago
@awvwgk I am very interested in this feature, is it still in the cards for this project? If so, I could take a swing at it, although I am unfamiliar with xtb and xtb-python source codes. Do you have some estimate of how much work would it be to add this?
Thanks!
Daniel Hollas University of Bristol
This would be the best issue to start looking into: https://github.com/grimme-lab/xtb/issues/403
Hello I would like to request an optimization scheme in the python module like the xTB program itself. I know this is possible with ASE, but ASE tends to slow down things. Currently im getting a timing of 0.05 seconds for a simple linear water molecule with no angle using ASE and about 0.2 seconds for a carbon nanotube with 120 C atoms. I believe if an optimization is done without using ASE, it can run faster. The reason for my request is that I am currently developing a VR platform for molecular simulation (scripting is done using python) and am implementing an interactive forcefield that needs a 5 step optimization to be run in approximately 0.015 seconds. If the optimization is slower, the relaxation will not be in "real-time". I can imagine it is possible as the xTB program already has the ability to optimize using the different methods.
Thank you Kind regards Zan Mahmood