How to analyse LST-1 data
Get the data
From the LST wiki page, or the elogs, find the runs you want to analyse. Find the corresponding DL1 files on the cluster, probably under /fefs/aswg/data/real/DL1/.../
. These files have been automatically produced by LSTOSA, there might be several versions, corresponding to different lstchain versions. In doubt, use the last one.
Select the data
You might want to sub-select the runs to use, following https://indico.cta-observatory.org/event/3984/
Corresponding MC production
The AllSky MC production follows sources declinations, as explained in: - https://indico.cta-observatory.org/event/4061/contributions/33409/attachments/21211/29956/Crab_analysis_20220404.pdf - ??
Random forests models are trained per declination and then used on the MC test data to produce IRFs on all test pointing directions.
Produce your DL2 files
From the DL1 files, you will need to produce the DL2 files using the right RF model.
Standard RF models and IRFs
If you are analysing an extragalactic source, the standard models trained for the closest declination of your source is likely enough. You can find these models under
/fefs/aswg/data/models/AllSky/.../
Apply the model to the DL1 files using lstchain lstchain_dl1_to_dl2
In this case, you can also use the corresponding standard IRFSs, under /fefs/aswg/data/IRF/AllSky/.../
.
Custom RF models and IRFs
If your source needs MC tuning (e.g. for strong NSB / galactic sources): 1. please check that a fitting custom training does not exist from https://cta-observatory.github.io/lstmcpipe/productions.html 2. if not, you may request a custom training using lstmcpipe pull-requests: https://cta-observatory.github.io/lstmcpipe/index.html#requesting-a-mc-analysis
The lstchain config must be produced using lstchain_tune_nsb
on the DL1 files you want to analyse.
The lstmcpipe config must be produced following: https://cta-observatory.github.io/lstmcpipe/pipeline.html#allsky-production-pipeline
Produce your DL3 files
lstchain_create_dl3_file
High-level analysis
Use gammapy to analyse the DL3 files using the corresponding IRFs.