Correlated uncertainty for new BM trees

[sevyharris]

Motivation or Problem

RMG does not currently have a way to estimate correlated uncertainties of reaction rates. The uncertainty tool by Gao et al is not compatible with the new BM trees and the new BM trees only have uncorrelated uncertainty estimates. This PR adds code that allows users to estimate correlated uncertainties.

Description of Changes

• Wrote functions to compute sensitivities of BM tree nodes to training reaction parameters
• Updated uncertainty.py to handle BM tree nodes as a reaction source
• Added example script for exporting mechanism correlated uncertainty

Testing

TODO

Reviewer Tips

TODO

Other TODO

• [ ] Make chemkin word wrap optional
• [ ] Add script for exporting sensitivities to RMG-database files
• [ ] Move export_corr_uncertainty to examples? Change it to a Jupyter Notebook?

[lgtm-com[bot]]

This pull request introduces 1 alert when merging 86da5fa8b4865505ac97a2ef66d300279475b9a0 into 517b347e183bfe0cdb8fb5655e859827ade25750 - view on LGTM.com

new alerts:

• 1 for Unused local variable

[codecov[bot]]

Codecov Report

Merging #2325 (a7f2510) into main (2cdc2ea) will increase coverage by 0.24%. The diff coverage is 2.85%.

:exclamation: Current head a7f2510 differs from pull request most recent head c835bad. Consider uploading reports for the commit c835bad to get more accurate results

@@            Coverage Diff             @@
##             main    #2325      +/-   ##
==========================================
+ Coverage   48.38%   48.62%   +0.24%     
==========================================
  Files         110      109       -1     
  Lines       30517    30198     -319     
  Branches     7951     7888      -63     
==========================================
- Hits        14765    14685      -80     
+ Misses      14211    13973     -238     
+ Partials     1541     1540       -1     

Impacted Files	Coverage Δ
rmgpy/data/kinetics/family.py	46.94% <2.85%> (-1.17%)	:arrow_down:
arkane/encorr/corr.py	41.33% <0.00%> (-21.34%)	:arrow_down:
arkane/encorr/decomp.py	93.50% <0.00%> (-5.20%)	:arrow_down:
arkane/ess/molpro.py	58.39% <0.00%> (-4.20%)	:arrow_down:
rmgpy/data/kinetics/database.py	50.36% <0.00%> (-1.08%)	:arrow_down:
rmgpy/molecule/fragment.py	73.33% <0.00%> (-0.30%)	:arrow_down:
rmgpy/rmg/pdep.py	12.05% <0.00%> (-0.29%)	:arrow_down:
arkane/pdep.py	25.33% <0.00%> (-0.29%)	:arrow_down:
rmgpy/data/thermo.py	66.10% <0.00%> (-0.27%)	:arrow_down:
arkane/input.py	64.19% <0.00%> (ø)
... and 5 more

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more

[lgtm-com[bot]]

This pull request introduces 1 alert when merging 94f704d10ffb9e0f8ea8fcd6513db137eabd840c into b2c917fd775c10e43bcb55805c9652ce05044532 - view on LGTM.com

new alerts:

• 1 for Unused local variable

[lgtm-com[bot]]

This pull request introduces 1 alert when merging b6d50f03d25f7b33c7e57b1524e2d07cc85f134f into 819779201aa4030c9b70c5ba6a39e929012161bb - view on LGTM.com

new alerts:

• 1 for Unused local variable

[lgtm-com[bot]]

This pull request introduces 1 alert when merging 427efdff7fc1b9f0515317988e89c4867a504e2d into 0df21b99f06cf7b2e1cf78ecdd5939af83706deb - view on LGTM.com

new alerts:

• 1 for Unused local variable

[github-actions[bot]]

This pull request is being automatically marked as stale because it has not received any interaction in the last 90 days. Please leave a comment if this is still a relevant pull request, otherwise it will automatically be closed in 30 days.