search

icecube / event-generator

IceCube event generator
GNU General Public License v3.0
1 stars 2 forks source link

Bug fix/time window exclusions #8

Closed mhuen closed 3 months ago

mhuen commented 1 year ago

Previously, when excluding time window exclusions, the individual PDFs of each mixture model component were re-normalized individually. This, however, is incorrect as this will modify the shape of the combined PDF. Instead, the entire mixture PDF has to be re-normalized by the total dom_cdf_exclusions_sum.

Note: this bugfix is incompatible with models trained in previous versions. To indicate this, the version number is increased.

codecov-commenter commented 3 months ago

Codecov Report

Attention: Patch coverage is 0% with 57 lines in your changes are missing coverage. Please review.

Please upload report for BASE (CollectBreakingChanges@769de28). Learn more about missing BASE report.

Files Patch % Lines
...manager/reconstruction/modules/visualize_pulses.py 0.00% 15 Missing :warning:
egenerator/ic3/visualization.py 0.00% 11 Missing :warning:
egenerator/model/source/noise/default.py 0.00% 9 Missing :warning:
egenerator/model/source/base.py 0.00% 8 Missing :warning:
egenerator/loss/default.py 0.00% 6 Missing :warning:
egenerator/model/multi_source/base.py 0.00% 4 Missing :warning:
egenerator/model/source/cascade/default.py 0.00% 4 Missing :warning:
Additional details and impacted files ```diff @@ Coverage Diff @@ ## CollectBreakingChanges #8 +/- ## ========================================================= Coverage ? 20.58% ========================================================= Files ? 87 Lines ? 9482 Branches ? 1727 ========================================================= Hits ? 1952 Misses ? 7413 Partials ? 117 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

mhuen commented 3 months ago

Note on the failed model prediction test: this test is expected to fail at the moment since an incompatibility check was introduced that triggers when trying to run with models of older software versions. Once all breaking changes are included in the branch CollectBreakingChanges, I will train a new model and update the test data to fix these unit tests.