CIF files with space group other than P1 are not correctly loaded
Opened this issue · 7 comments
Avogadro version: (please complete the following information from the About box):
- Avogadrolibs: 1.99.0
- Qt: 5.15.8
Desktop version: (please complete the following information):
- OS: macOS 14.3.1
- Version: 1.99.0
Describe the bug
CIF file with non-P1 symmetry is incorrectly loaded.
To Reproduce
Steps to reproduce the behavior:
- Download this valid CIF file https://gist.github.com/fxcoudert/71f75c57c2f86024e19eca12be39f92c
- Load it into Avogadro
- Observe a very partial structure (see below for the screenshot).
Expected behavior
The structure should show 576 atoms.
Screenshots
This happens even with the CIF files inside Avogadro itself:
It's loaded just fine. It's showing the primitive unit cell. If you want the whole thing, go to Crystal => Fill Unit Cell
(Looks like the French translation needs some work unfortunately.)
I welcome heuristics for knowing when people want the primitive vs. a filled cell. We had one in Avogadro 1.2 and lots of complaints.
Happy to continue the discussion: https://discuss.avogadro.cc/
But it thinks the system is P1:
And if I fill the cell, then it can't find the space group anymore:
I can change the tolerance but it still fails, even though it filled the cell itself.
Something else I tried: make a supercell (but that only duplicates the asymmetric unit). If you make a supercell THEN fill the cell, it generates something plain wrong.
Did you "perceive space group" before filling the unit cell? I did "fill unit cell" and it worked fine. If I did "perceive space group" just now, for some reason, it's saying "P1" which is for sure a bug (e.g., if the space group is already set, don't override it.)
What I don't understand is: it definitely read the space group and associated information, but it doesn't seem to display it anywhere? I can't find it.
Also, it's allowing to edit cell parameters inconsistently with the space group.
This issue has been mentioned on Avogadro Discussion. There might be relevant details there:
https://discuss.avogadro.cc/t/heuristic-for-reading-cif-crystal-structures-v2/5340/1