search

UCL-CCS / TIES

Topology Superimposition based on joint graph traversal
MIT License
5 stars 1 forks source link

use_element_in_superimposition #329

Closed AlexW12344 closed 1 year ago

AlexW12344 commented 1 year ago

Hiya,

I noticed that use_element_in_superimposition=True raises the warning 'Warning. Ignoring the specific atom types during superimposition. The results should not be used in production simulations.' but the description does not sound so bad https://github.com/UCL-CCS/TIES/blob/master/ties/pair.py#L90C9-L91C85 suggesting the atoms types are still considered in the final solution?

Should use_element_in_superimposition=True really be avoided for production. How might this effect the results?

Cheers, Alex

bieniekmateusz commented 1 year ago

Hey,

Oh yes that comment is an opinion more than anything else. This will simply narrow down the search. Sometimes it is nice to do the full "shape overlap" before trimming down the superimposition using the atom types. This might occasionally lead to a nicer overlap. I think in most of the cases it does not matter. I've been using an element almost always.

Best, Mat

AlexW12344 commented 1 year ago

Okay phew because I was using use_element_in_superimposition=True everywhere then saw this warning haha

bieniekmateusz commented 1 year ago

Sorry! I promise to clean it up a bit (next year it seems!)