search

SST-1M-collaboration / sst1mpipe

Single-Mirror Small Size Telescope for observations of gamma-rays above 3 TeV
http://sst-1m.space
BSD 3-Clause "New" or "Revised" License
0 stars 0 forks source link

Event ids are not matching #23

Open BastienLacave opened 2 months ago

BastienLacave commented 2 months ago

In the DL1 stereo files (maybe also for mono) of the daily analysis, the event ids in /dl1/event/telescope/parameters/tel_xxx and dl1/event/subarray/trigger are not matching.

In /dl1/event/telescope/parameters/tel_xxx, all event ids are unique, and I believe they are correct. In dl1/event/subarray/trigger, less than 12% are found in /dl1/event/telescope/parameters/tel_xxx, and there are duplicates of the same event id (even triplets), I believe this is not expected.

I was looking into this as CTLearn uses the event ids from the trigger, preventing me from matching theme with the DL2 events of the RF method.

tt-thomas commented 2 months ago

There was recent changes in how the event IDs are handled, since the update of the SWAT. Can you precise the dates of the run in which you had this issue, and the pipeline version used to process the dl1 data?

BastienLacave commented 2 months ago

this one is from 29/01/2024. I am not sure about the version of the pipeline however

jurysek commented 2 months ago

Yes, this is very dependent on the data file you are processing and on the pipeline version. From 20240130 SWAT has been tagging coincident events and arrayEvtNum is used as event_id. This particular event_id is used for stereo event matching in sst1mpipe >=0.3.3. Before that date and version, I cannot guarantee what event_id is in the /trigger, because we did not care too much.

I have just checked on some random data file from 20240213 using sst1mpipe v0.3.3, and I can confirm that event ids match in both tables (parameters and trigger) for both mono and stereo files.

But if this is still an issue in 0.4.1 run on data older than 20240130, please let us know and we can debug it.