search

IRL2 / nanover-protocol

https://irl2.github.io/nanover-docs/#
Other
0 stars 1 forks source link

Document explicitly (in README?) that ASE and OMM give different physics (integrators etc) #142

Open Ragzouken opened 2 months ago

Ragzouken commented 2 months ago

Reported for nanotube in SubtleGame

rhoslynroebuck commented 2 months ago

From testing, the nanotube looks the same (looking and interacting with it in VR) for the nanover.omm.server and the Rust server, but the physics looks different when using nanover.omm.ase.

Videos here, the issue might not be obvious just by looking at them:

https://github.com/IRL2/nanover-protocol/assets/50617818/81522681-602d-49e4-b848-c261984f3f13

https://github.com/IRL2/nanover-protocol/assets/50617818/c558718b-3930-46a5-a77c-7c3dc582e466

https://github.com/IRL2/nanover-protocol/assets/50617818/d8c96b03-bd4a-4fba-93aa-c7415fc9ffd4

phuddha commented 2 months ago

the issue is with open-mm vs ase (with the python server) - hence changing name of issue. ASE = Atomic Simulation Environment (https://wiki.fysik.dtu.dk/ase/about.html). My (potentially naive) understanding is that Nanover supports ASE based energy / force calculations, minimisations etc (vs openmm native capabilities). @mcsloy @hjstroud - isn't different physics actually expected in this case?

jbarnoud commented 2 months ago

The physics are expected to be different between OpenMM itself and OpenMM with ASE. In the latter case, the integrator is the ASE one rather than the OpenMM one. In addition, Nanover omm ase does not read the integrator parameters from the XML input.

Ragzouken commented 2 months ago

The physics are expected to be different between OpenMM itself and OpenMM with ASE. In the latter case, the integrator is the ASE one rather than the OpenMM one. In addition, Nanover omm ase does not read the integrator parameters from the XML input.

Are those parameters applicable to ASE, would it make sense to be reading them?

jbarnoud commented 2 months ago

You could make some cases match roughly, but the XML can define integrators that are not in ase and those who are will still have discrepancies.

On 10 June 2024 13:10:14 CEST, mark wonnacott @.***> wrote:

The physics are expected to be different between OpenMM itself and OpenMM with ASE. In the latter case, the integrator is the ASE one rather than the OpenMM one. In addition, Nanover omm ase does not read the integrator parameters from the XML input.

Are those parameters applicable to ASE, would it make sense to be reading them?

-- Reply to this email directly or view it on GitHub: https://github.com/IRL2/nanover-protocol/issues/142#issuecomment-2158052912 You are receiving this because you commented.

Message ID: @.***>