Lester Hedges
Lester Hedges
No problem. The post magically updated just as I hit comment :-)
We are using the _exact_ same reduction scheme as OpenMM. The issue is that their C++ API uses _exact_ numerical precision in their checks, whereas their Python API uses a...
Could you try the attempted "fix" that I've commited [here](https://github.com/OpenBioSim/sire/commit/e7c54e9d6b50bbd51030be23c5c81b9473bad94c), which is on the [fix_305](https://github.com/OpenBioSim/sire/tree/fix_305) branch This reapplies the reduction on the triclinic space object just before the box vectors...
Any update on this? If it works, I'll merge across to my other fix branch and get this into `devel` as soon as possible. If you're unable to test, can...
Yes, that's right. As mentioned up thread, I only applied this to the regular, non-PME, version of the code. If it works, I'll patch it into all SOMD variants.
Thanks, I'll debug locally to see what's going on.
It seems to be okay if you only reduce the box, not rotate it. Could you try removing the line: ```cpp space.rotate(); ``` I assume that we are only having...
For the rotation, the issue is likely that the center of rotation is wrong given the current box vectors and origin. This can be specified when using the BioSimSpace Python...
Great, thanks for confirming. I'll add this to all SOMD flavours and create a PR.
Many thanks. I've resolved this locally but hadn't pushed a fix (mostly because it doesn't affect the Python versions used for the CI). @chryswoods: I think the fix on your...