PyGRB reweighted SNR cut bug

[pannarale]

@Thomas-JACQUOT uncovered two bugs when running PyGRB without vetoes.

First bug

pycbc_pygrb_plot_injs_results was not applying the reweighted SNR cut to the reweighted SNR values of found injections. This would then lead to indexing errors such as

IndexError: boolean index did not match indexed array along dimension 0; dimension is 6080 but corresponding boolean dimension is 5827

This PR adds one line to pycbc_pygrb_plot_injs_results to get it to run correctly, even when vetoes are not used.

Second bug

pycbc_pygrb_plot_null_stats also does not work correctly when using it without vetoes. Namely, it does not find the key for coincident SNR. This happens because these values are calculated upon the user's request, in an if statement that checks whether vetos are being used.

This PR fixes the function apply_vetoes_to_found_injs in pycbc.results.pygrb_postprocessing_utils.py so that regardless of whether the user passes vetoed times or not, the same final quantities to be plotted are calculated and stored.

This PR only affects PyGRB and fixes both problems.

I encourage @Thomas-JACQUOT to use this branch to ensure that his run goes to completion.

• The author of this pull request confirms they will adhere to the code of conduct

[pannarale]

For reference, I am posting the output of two pycbc_pygrb_plot_injs_results jobs, with and without the veto-file

(o4_clean_master) [francesco.pannarale@ldas-osg2 post_processing]$ pycbc_pygrb_plot_injs_results --verbose  --newsnr-threshold 6. --trig-file H1L1V1-PYGRB_GRB170817A_OFFSOURCE_CLUSTERED-1187006058-5648.h5 --seg-files bufferSeg.txt offSourceSeg.txt onSourceSeg.txt  --trigger-time 1187008884 --found-missed-file HL-INJECTIONS_SPLIT_INSPINJ_NSNS_FOUNDMISSED-1187006058-5648.h5 --output-file H1L1V1-PYGRB_PLOT_INJS_RESULTS_MCHIRP_INCL_FOUND-ON-TOP_NSNS_GRB170817A-1187006058-5648.png --y-variable mchirp --x-variable incl
2025-01-24T08:01:33.613-08:00: INFO: Setting output directory.
2025-01-24T08:01:33.614-08:00: INFO: Loading timeslides.
2025-01-24T08:01:33.619-08:00: INFO: Loading segments.
2025-01-24T08:01:33.620-08:00: INFO: Constructing trials.
2025-01-24T08:01:33.671-08:00: INFO: 9834 trials generated.
2025-01-24T08:01:33.674-08:00: INFO: 17980 triggers loaded.
2025-01-24T08:01:33.679-08:00: INFO: 15455 triggers surviving reweighted SNR cut at 6.0.
2025-01-24T08:01:34.491-08:00: INFO: 15273 triggers found within the trials dictionary and sorted.
2025-01-24T08:01:34.499-08:00: INFO: Triggers information extracted.
2025-01-24T08:01:34.560-08:00: INFO: Background bestNR calculated.
2025-01-24T08:01:34.561-08:00: INFO: 1103 injections loaded.
2025-01-24T08:01:34.562-08:00: INFO: 1103 injections surviving reweighted SNR cut at None.
2025-01-24T08:01:34.622-08:00: INFO: 1397 missed injections analysed.
2025-01-24T08:01:34.622-08:00: INFO: 1103 found injections analysed.
2025-01-24T08:01:34.623-08:00: INFO: After applying reweighted SNR cut at 6.0: 1098 found injections and 1402 missed injections
2025-01-24T08:01:35.032-08:00: INFO: Plot complete.
(o4_clean_master) [francesco.pannarale@ldas-osg2 post_processing]$ pycbc_pygrb_plot_injs_results --verbose  --newsnr-threshold 6. --trig-file H1L1V1-PYGRB_GRB170817A_OFFSOURCE_CLUSTERED-1187006058-5648.h5 --seg-files bufferSeg.txt offSourceSeg.txt onSourceSeg.txt  --trigger-time 1187008884 --found-missed-file HL-INJECTIONS_SPLIT_INSPINJ_NSNS_FOUNDMISSED-1187006058-5648.h5 --output-file H1L1V1-PYGRB_PLOT_INJS_RESULTS_MCHIRP_INCL_FOUND-ON-TOP_NSNS_GRB170817A-1187006058-5648.png --y-variable mchirp --x-variable incl --veto-file H1L1V1-VETOES-1187006179-5502.xml
2025-01-24T08:01:44.487-08:00: INFO: Setting output directory.
2025-01-24T08:01:44.488-08:00: INFO: Loading timeslides.
2025-01-24T08:01:44.493-08:00: INFO: Loading segments.
2025-01-24T08:01:44.494-08:00: INFO: Constructing trials.
2025-01-24T08:01:44.506-08:00: INFO: 121s of data vetoed at GPS time 1187006058
2025-01-24T08:01:44.506-08:00: INFO: 2s of data vetoed at GPS time 1187009326
2025-01-24T08:01:44.506-08:00: INFO: 2s of data vetoed at GPS time 1187010566
2025-01-24T08:01:44.506-08:00: INFO: 25s of data vetoed at GPS time 1187011681
2025-01-24T08:01:44.506-08:00: INFO: 121s of data vetoed at GPS time 1187006058
2025-01-24T08:01:44.507-08:00: INFO: 25s of data vetoed at GPS time 1187011681
2025-01-24T08:01:44.507-08:00: INFO: 121s of data vetoed at GPS time 1187006058
2025-01-24T08:01:44.507-08:00: INFO: 25s of data vetoed at GPS time 1187011681
2025-01-24T08:01:44.572-08:00: INFO: 9490 trials generated.
2025-01-24T08:01:44.575-08:00: INFO: 17980 triggers loaded.
2025-01-24T08:01:44.578-08:00: INFO: 15455 triggers surviving reweighted SNR cut at 6.0.
2025-01-24T08:01:45.397-08:00: INFO: 15171 triggers found within the trials dictionary and sorted.
2025-01-24T08:01:45.405-08:00: INFO: Triggers information extracted.
2025-01-24T08:01:45.464-08:00: INFO: Background bestNR calculated.
2025-01-24T08:01:45.465-08:00: INFO: 1103 injections loaded.
2025-01-24T08:01:45.466-08:00: INFO: 1103 injections surviving reweighted SNR cut at None.
2025-01-24T08:01:45.526-08:00: INFO: 1397 missed injections analysed.
2025-01-24T08:01:45.526-08:00: INFO: 1103 found injections analysed.
2025-01-24T08:01:45.526-08:00: INFO: After applying reweighted SNR cut at 6.0: 1098 found injections and 1402 missed injections
2025-01-24T08:01:45.526-08:00: INFO: Applying data vetoes to found injections...
2025-01-24T08:01:45.530-08:00: INFO: 0 injections vetoed due to H1.
2025-01-24T08:01:45.538-08:00: INFO: 0 injections vetoed due to L1.
2025-01-24T08:01:45.545-08:00: INFO: 0 injections vetoed due to V1.
2025-01-24T08:01:45.549-08:00: INFO: 0 injections vetoed.
2025-01-24T08:01:45.549-08:00: INFO: 1098 injections surviving vetoes.
2025-01-24T08:01:45.940-08:00: INFO: Plot complete.

[Thomas-JACQUOT]

I've tested this modification on my jobs and it worked ! 🎉

[pannarale]

@Thomas-JACQUOT's comment above refers to pycbc_pygrb_plot_injs_results.