Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option #4549

GarethCabournDavies · 2023-10-31T12:11:05Z

When we have lots of things in the autogating vetoes, these can dominate the singles minifollowups.

This allows us to provide the gating vetoes to this job as elsewhere in order to remove them.

However they are still informative (e.g. we may have an autogated loud signal!). So we also have a minifollowup with --vetoed-time-only option.

The main changes here are:

Add the gating vetoes options and implement
Allow a --vetoed-time-only option
Allow passing of both veto file and foreground-censor file, this is as we do not want foreground-censored things to show up in a 'vetoed time only' section
- Update the offline workflow to use this specification as appropriate

… vetoes' option

GarethCabournDavies · 2023-11-01T09:42:41Z

(I've added the v23 release branch label, but this is not vital, more a helpful addition)

GarethCabournDavies · 2023-11-01T14:02:11Z

An example (behind LVK paywall) with the O4a chunk 2 test is here

maxtrevor

Looks good to me, approving

tdent · 2023-11-02T15:33:44Z

It looks like it works, but I am not quite convinced that this will be what we want to do, ie will it allow us to catch a moderately loud signal whose merger happens to lie in a gating veto.
The 'loudest vetoed' sections in the example (in both ifos) list things which even at event number 10 are still pretty loud, and they all have really high masses. If there was (say) a moderately loud low mass signal with a merger 1-2s after a gating window, it might not make it to this list.

tdent · 2023-11-02T15:36:48Z

Maybe what would mitigate this would be to apply some more intelligent statistic ranking (ie loudest by what statistic - IMO it should ideally be the one used for detecting singles) to the 'loudest vetoed' section ..

Anyway I will ask for the release branch tag to be removed until we have thought about this a bit more.

GarethCabournDavies · 2023-11-02T15:41:10Z

That's a fair point, and I think having the ranking-statistic option with quadsum as default (i.e. pass sngl-ranking straight through), and calculating the single-detector statistic should work for that. I've removed the release branch tag and will remove the request on the cherry-pick PR as well

tdent · 2023-11-02T15:45:35Z

OK, so your example used newsnr_sgveto_psdvar_threshold ranking .. maybe activating the actual sngls detection statistic needs a bit more code plumbing then.

tdent · 2023-11-03T13:22:10Z

FWIW, I don't think there is a reason to not merge this now, unless you want to mark it as draft and add more actual code changes later.

GarethCabournDavies · 2023-11-03T13:24:37Z

I am already very close to getting the ranking-statistic changes in, so it may be worth waiting, though I will mark as draft for now

GarethCabournDavies · 2023-11-07T09:38:54Z

So I was completely done when I said 'very close' before. But it turned out there was a bug when calling rank_stat_single with the quadsum statistic, this took me longer than I care to admit to find, so the tests were all failing. Workflow running now, and I will link it and mark as ready when it is done

GarethCabournDavies · 2023-11-09T12:54:44Z

A couple of minor associated fixes, and we have results again:

https://ldas-jobs.ligo.caltech.edu/~gareth.cabourndavies/testoutput/veto_sngl_minifollowup/a2_TEST_2/

GarethCabournDavies · 2023-11-09T13:09:36Z

See GarethCabournDavies/pycbc@c53e31d...sngls_minifollow_gating_veto for changes since the last review

GarethCabournDavies · 2023-11-09T16:41:11Z

I think this is ready now - I'll be away next week, so please merge if we are happy with it, and apply the release branch tag if we are going to use it in O4 production runs

GarethCabournDavies · 2023-11-16T08:50:58Z

Any comments since last review? If not I will merge

GarethCabournDavies · 2023-11-23T13:32:57Z

No comments against this, so I will merge

…toes' option (gwastro#4549) * Use gating windows in sngl_minifollowup - also allow 'triggers within vetoes' option * Minor fixes * add infor if fewer triggers than requested * Allow pycbc_sngls_minifollowup to use single ranking statistic * Fix bug in rank_stat_single for quadsum/phasetd statistics * mask may be boolean now * fix broken pycbc_page_snglinfo due to bugfix * page_snglinfo to use the ranking statistic used * missed change

tdent · 2024-01-08T13:51:24Z

So did we agree that this is now not a risk for missing loudish signals with merger inside a gating veto window?

GarethCabournDavies · 2024-01-08T14:16:16Z

These would show up in the 'within vetoes' followup page, so long as that is specified in the config files

…toes' option (gwastro#4549) * Use gating windows in sngl_minifollowup - also allow 'triggers within vetoes' option * Minor fixes * add infor if fewer triggers than requested * Allow pycbc_sngls_minifollowup to use single ranking statistic * Fix bug in rank_stat_single for quadsum/phasetd statistics * mask may be boolean now * fix broken pycbc_page_snglinfo due to bugfix * page_snglinfo to use the ranking statistic used * missed change

tdent · 2024-01-08T14:28:06Z

OK, is a MR required for that?

GarethCabournDavies · 2024-01-08T14:43:59Z

Added one for the CI here: #4597

There is also one for O4 production runs in the repo for that

tdent · 2024-01-08T15:32:27Z

Ah, I see one there (currently in draft state I think)

* Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option (#4549) * Use gating windows in sngl_minifollowup - also allow 'triggers within vetoes' option * Minor fixes * add infor if fewer triggers than requested * Allow pycbc_sngls_minifollowup to use single ranking statistic * Fix bug in rank_stat_single for quadsum/phasetd statistics * mask may be boolean now * fix broken pycbc_page_snglinfo due to bugfix * page_snglinfo to use the ranking statistic used * missed change * Fix cases where ifar limit was not being applied (#4574) * Fix cases where ifar limit was not being applied * some more missed cases / long lines * redoing a couple more long lines * missed one more place * unrelated bug where the IFOs were not in the right order for use with the significance_dict * remove removal of downranked triggers (#4579) * remove removal of downranked triggers * add brief explanation into the caption * bug in choosing the far calculation method (#4593) * bug in which IFO combo is used for calculating the FAR in each combination during HR * Safety catch * Update release number * add coordinates_space.py (#4289) * add coordinates_space.py * add LISA/SSB frame params * add LISA_to_GEO and GEO_to_LISA * add coordinates_space into FieldArray * add doc and Astropy support * update comments on sympy * use fsolve from scipy instead * fix cc issues * fix cc issues * minor fix * update * not use iteration * decouple LISA orbit and more accurate Earth * rename * remove jplephem * add the angular displacement of the Earth * use radians * make func readable in .ini * reverse back to master * correct psi range * reverse to master * fix unit issue in earth_position_SSB * put LISA to the "right" position * add LISA specific transform classes here * change names * update * make a package for coordinates * remove coordinates_space import * move __all__ into __init__.py * remove all coordinates_space * change TIME_OFFSET to seconds * fix SOBHB issue * rename * add SSB or LISA params into fid_params * rename * fix cc issues * fix cc issue * fix cc issue * update * update * fix * add default names * overwrite params with same names * remove pre-fixed names * remove all pre-fixed names * not pop * fix inverse transform * update tc * not overwrite * add SNR support for multi-model * Update waveform.py * t0 issue * t0 issue * Update space.py * add obstime * np.mod(psi_newframe, 2*np.pi) * fix obstime * add support for array inputs * Update hierarchical.py * just use Alex's implementation * CustomTransformMultiOutputs is in another PR, so remove it * add LDC and LAL convention correction * use pycbc standard names * more meaningful name * use pycbc standard names * Update relbin.py * Update parameters.py * remove unnecessary changes * fix cc issue * fix cc issue * fix cc issue * fix cc issue * compactify * compactify * add __all__ back * Update transforms.py * Update transforms.py * Update test_transforms.py * Update transforms.py * update doc * fix time warning * Update space.py * Update test_transforms.py * Create test_coordinates_space.py * fix cc issues * fix cc issues * fix cc issue * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update test_coordinates_space.py * add inline doc * Update tox.ini * add check of bbhx * Update test_coordinates_space.py * Update tox.ini * Update test_coordinates_space.py * add MultibandRelativeTimeDom into hierarchical.py * Update __init__.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update hierarchical.py * Update hierarchical.py * Update hierarchical.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update __init__.py * Update space.py * Update space.py * Update space.py * fix psi issue * Update test_coordinates_space.py * Update test_coordinates_space.py * update lalsimulation cvmfs path (#4580) * update lalsimulation cvmfs path * missed that the mount location needs to be changed as well * more references to previsou CVMFS location * Revert "add coordinates_space.py (#4289)" This reverts commit e3418c7. * REmoving lisa examples * REmove inference examples * Remove LISA deps * Removing more LISA things --------- Co-authored-by: Gareth S Cabourn Davies <[email protected]> Co-authored-by: Shichao Wu <[email protected]>

tdent · 2024-01-09T17:46:34Z

Ultimately we should replace gating veto with re-ranking around a gated time, ie gating DQ segment .. that is under development

…toes' option (gwastro#4549) * Use gating windows in sngl_minifollowup - also allow 'triggers within vetoes' option * Minor fixes * add infor if fewer triggers than requested * Allow pycbc_sngls_minifollowup to use single ranking statistic * Fix bug in rank_stat_single for quadsum/phasetd statistics * mask may be boolean now * fix broken pycbc_page_snglinfo due to bugfix * page_snglinfo to use the ranking statistic used * missed change

* Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option (gwastro#4549) * Use gating windows in sngl_minifollowup - also allow 'triggers within vetoes' option * Minor fixes * add infor if fewer triggers than requested * Allow pycbc_sngls_minifollowup to use single ranking statistic * Fix bug in rank_stat_single for quadsum/phasetd statistics * mask may be boolean now * fix broken pycbc_page_snglinfo due to bugfix * page_snglinfo to use the ranking statistic used * missed change * Fix cases where ifar limit was not being applied (gwastro#4574) * Fix cases where ifar limit was not being applied * some more missed cases / long lines * redoing a couple more long lines * missed one more place * unrelated bug where the IFOs were not in the right order for use with the significance_dict * remove removal of downranked triggers (gwastro#4579) * remove removal of downranked triggers * add brief explanation into the caption * bug in choosing the far calculation method (gwastro#4593) * bug in which IFO combo is used for calculating the FAR in each combination during HR * Safety catch * Update release number * add coordinates_space.py (gwastro#4289) * add coordinates_space.py * add LISA/SSB frame params * add LISA_to_GEO and GEO_to_LISA * add coordinates_space into FieldArray * add doc and Astropy support * update comments on sympy * use fsolve from scipy instead * fix cc issues * fix cc issues * minor fix * update * not use iteration * decouple LISA orbit and more accurate Earth * rename * remove jplephem * add the angular displacement of the Earth * use radians * make func readable in .ini * reverse back to master * correct psi range * reverse to master * fix unit issue in earth_position_SSB * put LISA to the "right" position * add LISA specific transform classes here * change names * update * make a package for coordinates * remove coordinates_space import * move __all__ into __init__.py * remove all coordinates_space * change TIME_OFFSET to seconds * fix SOBHB issue * rename * add SSB or LISA params into fid_params * rename * fix cc issues * fix cc issue * fix cc issue * update * update * fix * add default names * overwrite params with same names * remove pre-fixed names * remove all pre-fixed names * not pop * fix inverse transform * update tc * not overwrite * add SNR support for multi-model * Update waveform.py * t0 issue * t0 issue * Update space.py * add obstime * np.mod(psi_newframe, 2*np.pi) * fix obstime * add support for array inputs * Update hierarchical.py * just use Alex's implementation * CustomTransformMultiOutputs is in another PR, so remove it * add LDC and LAL convention correction * use pycbc standard names * more meaningful name * use pycbc standard names * Update relbin.py * Update parameters.py * remove unnecessary changes * fix cc issue * fix cc issue * fix cc issue * fix cc issue * compactify * compactify * add __all__ back * Update transforms.py * Update transforms.py * Update test_transforms.py * Update transforms.py * update doc * fix time warning * Update space.py * Update test_transforms.py * Create test_coordinates_space.py * fix cc issues * fix cc issues * fix cc issue * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update tox.ini * Update test_coordinates_space.py * add inline doc * Update tox.ini * add check of bbhx * Update test_coordinates_space.py * Update tox.ini * Update test_coordinates_space.py * add MultibandRelativeTimeDom into hierarchical.py * Update __init__.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update hierarchical.py * Update hierarchical.py * Update hierarchical.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update hierarchical.py * Update hierarchical.py * Update relbin.py * Update __init__.py * Update space.py * Update space.py * Update space.py * fix psi issue * Update test_coordinates_space.py * Update test_coordinates_space.py * update lalsimulation cvmfs path (gwastro#4580) * update lalsimulation cvmfs path * missed that the mount location needs to be changed as well * more references to previsou CVMFS location * Revert "add coordinates_space.py (gwastro#4289)" This reverts commit e3418c7. * REmoving lisa examples * REmove inference examples * Remove LISA deps * Removing more LISA things --------- Co-authored-by: Gareth S Cabourn Davies <[email protected]> Co-authored-by: Shichao Wu <[email protected]>

GarethCabournDavies added 3 commits October 31, 2023 04:42

Use gating windows in sngl_minifollowup - also allow 'triggers within…

269e993

… vetoes' option

Minor fixes

fcc6013

add infor if fewer triggers than requested

f96d653

GarethCabournDavies added the offline search label Oct 31, 2023

GarethCabournDavies requested a review from titodalcanton October 31, 2023 12:11

GarethCabournDavies added the v23_release_branch PRs applied to the v2.3.X release branch or to be cherry-picked if merging to master label Nov 1, 2023

GarethCabournDavies requested review from tdent, veronica-villa and maxtrevor November 2, 2023 09:34

GarethCabournDavies mentioned this pull request Nov 2, 2023

Cherry pick things onto v23 release branch #4551

Merged

tdent removed the request for review from titodalcanton November 2, 2023 14:52

maxtrevor approved these changes Nov 2, 2023

View reviewed changes

GarethCabournDavies removed the v23_release_branch PRs applied to the v2.3.X release branch or to be cherry-picked if merging to master label Nov 2, 2023

GarethCabournDavies marked this pull request as draft November 3, 2023 13:24

GarethCabournDavies added 3 commits November 7, 2023 09:17

Merge branch 'gwastro:master' into sngls_minifollow_gating_veto

c53e31d

Allow pycbc_sngls_minifollowup to use single ranking statistic

9c7c79f

Fix bug in rank_stat_single for quadsum/phasetd statistics

85df495

GarethCabournDavies added 3 commits November 7, 2023 05:06

mask may be boolean now

0ae6897

fix broken pycbc_page_snglinfo due to bugfix

54a7865

page_snglinfo to use the ranking statistic used

2973791

GarethCabournDavies marked this pull request as ready for review November 9, 2023 12:59

missed change

771d12a

GarethCabournDavies requested a review from maxtrevor November 9, 2023 16:52

GarethCabournDavies added the v23_release_branch PRs applied to the v2.3.X release branch or to be cherry-picked if merging to master label Nov 16, 2023

GarethCabournDavies merged commit a203541 into gwastro:master Nov 23, 2023
31 checks passed

GarethCabournDavies deleted the sngls_minifollow_gating_veto branch November 23, 2023 13:33

GarethCabournDavies mentioned this pull request Jan 8, 2024

Use vetoed times followup in example #4597

Merged

spxiwh removed the v23_release_branch PRs applied to the v2.3.X release branch or to be cherry-picked if merging to master label Jan 16, 2024

This was referenced Jul 24, 2024

Fix sngls_minifollowup event selection/ordering [master] #4823

Merged

Fix sngls_minifollowup event selection/ordering [v23_release_branch] #4824

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option #4549

Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option #4549

GarethCabournDavies commented Oct 31, 2023

GarethCabournDavies commented Nov 1, 2023

GarethCabournDavies commented Nov 1, 2023

maxtrevor left a comment

tdent commented Nov 2, 2023

tdent commented Nov 2, 2023 •

edited

Loading

GarethCabournDavies commented Nov 2, 2023

tdent commented Nov 2, 2023

tdent commented Nov 3, 2023

GarethCabournDavies commented Nov 3, 2023

GarethCabournDavies commented Nov 7, 2023

GarethCabournDavies commented Nov 9, 2023

GarethCabournDavies commented Nov 9, 2023 •

edited

Loading

GarethCabournDavies commented Nov 9, 2023

GarethCabournDavies commented Nov 16, 2023

GarethCabournDavies commented Nov 23, 2023

tdent commented Jan 8, 2024

GarethCabournDavies commented Jan 8, 2024

tdent commented Jan 8, 2024

GarethCabournDavies commented Jan 8, 2024

tdent commented Jan 8, 2024

tdent commented Jan 9, 2024

Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option #4549

Use gating windows in sngl_minifollowup and allow 'triggers within vetoes' option #4549

Conversation

GarethCabournDavies commented Oct 31, 2023

GarethCabournDavies commented Nov 1, 2023

GarethCabournDavies commented Nov 1, 2023

maxtrevor left a comment

Choose a reason for hiding this comment

tdent commented Nov 2, 2023

tdent commented Nov 2, 2023 • edited Loading

GarethCabournDavies commented Nov 2, 2023

tdent commented Nov 2, 2023

tdent commented Nov 3, 2023

GarethCabournDavies commented Nov 3, 2023

GarethCabournDavies commented Nov 7, 2023

GarethCabournDavies commented Nov 9, 2023

GarethCabournDavies commented Nov 9, 2023 • edited Loading

GarethCabournDavies commented Nov 9, 2023

GarethCabournDavies commented Nov 16, 2023

GarethCabournDavies commented Nov 23, 2023

tdent commented Jan 8, 2024

GarethCabournDavies commented Jan 8, 2024

tdent commented Jan 8, 2024

GarethCabournDavies commented Jan 8, 2024

tdent commented Jan 8, 2024

tdent commented Jan 9, 2024

tdent commented Nov 2, 2023 •

edited

Loading

GarethCabournDavies commented Nov 9, 2023 •

edited

Loading