Update LobsterTaskDoc

[naik-aakash]

Summary

The changes in the are first step toward LOBSTER dataset available via MP

Todo

• [x] decouple file pymatgen file parsing and post-processing (lobsterpy)
• [x] Set save_computational_data_jsons and add_coxxcar_to_task_document default to false
• [x] Adapt tests to accommodate the update in defaults
• [x] MoveBandoverlaps ,Charge, Grosspop, Icohplist, MadelungEnergies, SitePotential fields to data (Objects could be larger than 16MB for large structure)
• [x] Option to manipulate lobsterpy kwargs
• [x] redefine fields to have minimum redundant information
• [x] Move Charge, MadelungEnergies back to mongostore (size we suspect will not be larger than 16 MB , as these are simple text files with minimal data in them)
• [x] Nest strongest_bonds fields

[naik-aakash]

Hi @janosh , @utf , any ideas how to fix this issue properly for this pydantic error ? https://github.com/materialsproject/atomate2/pull/723/commits/5c88c8fcd04bf4cfc2c2fe8ee7c372c5e32a0c59

Do I need to define again new Models here for all this objects I made MSONable in pymatgen? Or simply pass them to datastore ? https://github.com/materialsproject/pymatgen/pull/3627

[naik-aakash]

Hi @janosh , @utf , any ideas how to fix this issue properly for this pydantic error ? 5c88c8f

Do I need to define again new Models here for all this objects I made MSONable in pymatgen? Or simply pass them to datastore ? materialsproject/pymatgen#3627

Sorry, nevermind, I think I got the issue. Just a new release of pymatgen should fix it I think. I had installed pymatgen from repo locally and tests passed and I forgot this for a moment 😅

[naik-aakash]

Hi @munrojm, can you please have a look at the updated schema? We have now made it possible to turn off the LobsterPy analysis . So this can be used to just parse the outputs.

Let me know if anything needs to be changed. If it is fine, maybe you can run some tests with it.

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 90.00000% with 5 lines in your changes are missing coverage. Please review.

Project coverage is 74.86%. Comparing base (256b39a) to head (81e1353). Report is 15 commits behind head on main.

:exclamation: Current head 81e1353 differs from pull request most recent head d3eb63a. Consider uploading reports for the commit d3eb63a to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #723      +/-   ##
==========================================
- Coverage   77.03%   74.86%   -2.17%     
==========================================
  Files         122      136      +14     
  Lines        9108    10519    +1411     
  Branches     1429     1644     +215     
==========================================
+ Hits         7016     7875     +859     
- Misses       1663     2154     +491     
- Partials      429      490      +61     

Files	Coverage Δ
src/atomate2/lobster/jobs.py	94.28% <100.00%> (ø)
src/atomate2/lobster/schemas.py	90.52% <89.58%> (-1.50%)	:arrow_down:

... and 2 files with indirect coverage changes

[munrojm]

@naik-aakash the updated doc with and without analysis still still extremely large when parsed. The one example I have is over 200MB uncompressed for both the former and latter. Looks like large volumetric data still should be factored out in the same way DOS and BS data is done in the other task documents.

[naik-aakash]

@naik-aakash the updated doc with and without analysis still still extremely large when parsed. The one example I have is over 200MB uncompressed for both the former and latter. Looks like large volumetric data still should be factored out in the same way DOS and BS data is done in the other task documents.

Hi @munrojm , Large objects are already factored out to datastore especially generated from Lobster. Am not sure I got what you meant to say. Can you please be more specific?

[JaGeo]

@munrojm it might also already help if you let us know which material you tested so that we can run tests ourselves. Thanks!

[naik-aakash]

Hi @utf , if you have any specific comments on the changes. Let us know . Or else it is ready to be merged.

[JaGeo]

@naik-aakash there are some conflicts here