Conversation
…/ml-peg into mandy_new_benchmark
ElliottKasoar
added
the
new benchmark
ElliottKasoar
left a comment
ElliottKasoar
left a comment
There was a problem hiding this comment.
Thank for this, @MandyHoffmann, it's looking really good!
Are you able to share (e.g. via slack/onedrive) some full results for this, so I can test the analysis/app more carefully?
Co-authored-by: Elliott Kasoar <45317199+ElliottKasoar@users.noreply.github.com>
…es.py Co-authored-by: Elliott Kasoar <45317199+ElliottKasoar@users.noreply.github.com>
…poles.py Co-authored-by: Elliott Kasoar <45317199+ElliottKasoar@users.noreply.github.com>
|
Hey @MandyHoffmann, i've made some code suggestions which adds the data downlaod. If you're happy with the change to janus then we could begin to clean up the code as well to get rid of the commented out sections. |
|
i also assume we can drop the changes to the |
Yes, @MandyHoffmann would you be able to revery the changes to the When you look at the files changed tab you only really want to see changes that are specific to your new benchmark. (In the terminal these changes would be shown in the |
Co-authored-by: Joseph Hart <92541539+joehart2001@users.noreply.github.com>
…poles.py DATA_PATH is not needed because of data download instead Co-authored-by: Joseph Hart <92541539+joehart2001@users.noreply.github.com>
…poles.py Co-authored-by: Joseph Hart <92541539+joehart2001@users.noreply.github.com>
|
Also one small thing, could you remove the data dir @MandyHoffmann as you’ve now added the download |
3 participants
Pre-review checklist for PR author
PR author must check the checkboxes below when creating the PR.
Summary
Physicality test for total dipole distribution of water slabs, checks standard deviation and number of candidates for dielectric breakdown.
Linked issue
Resolves #348
Progress
I think I did everything, but please check! For actually running it, the MD calculation time can be adjusted based on how well sampled you want it; longer simulation times are required for a proper Gaussian-shaped histogram. Similarly, one could consider changing the bins to be even smaller.
Currently, no hover templates are set for the histograms.
Testing
Tested on one of my own water models and the small mace-mp0-a model. Calculations are quite expensive (>12 h for the mace-mp0 one).
New decorators/callbacks
Build a new decorator for histograms, oriented at the scatter plot one, included in this PR. Not sure all the callbacks work properly, but one can see the plots.