Update some functions for DL3 Tool

[chaimain]

Update pyirf dependency, and some restructuring of functions relevant for DL3 Tool to accommodate future PRs

Functions changes -

1. Two functions in irf/hdu_table.py.

From create_event_list, create some separate functions -

• get_timing_params: Get a dict of timing parameters to be added in the DL3
• get_pointing_params: Get telescope pointing in ICRS, source pointing angular position from camera centre and the respective observation mode for DL3
• add_icrs_position_params: Transform the reconstructed position from Alt/Az system to RA/Dec and use astropy.erfa_astrom Interpolation for a fast transformation, and to also get the theta angle with respect to the source position provided. These parameters are added to the DL2 file as separate columns and the create_event_list uses it to fill the EVENTS HDU. NOTE: erfa_astrom interpolation method might need to be updated for an even quicker but also accurate process to transform the coordinates. Hence it is essential to separate this functionality, to be able to test more easily.

In create_hdu_index_hdu function, all IRF HDUs are added by iterating over a given list of HDU names. For energy-dependent cuts, 2 extra HDUs are added - "RAD_MAX" and "GH CUTS". "GH CUTS" HDU is a new temporary HDU being added in PR#709.

2. 1 function in reco/utils.py

To calculate the orthogonal geomagnetic angle (delta). The geomagnetic parameters are calculated using values for LST1 location in https://www.ngdc.noaa.gov/geomag/calculators/magcalc.shtml?#igrfwmm and by using the predicted change as per the IGRF model, for a given time between 2020-2025. This will be used in PR#711

This modification is necessary so that once the 2 PRs are merged, it will be smoother to integrate, and also for any future update on the DL3 Tools.

Also, 1 more filter on event selection for DL3 is added, to select only cosmic events (event_type=32), to be more accurate in our choice of event selection.

[codecov]

Codecov Report

Merging #752 (e67d553) into master (f49becb) will decrease coverage by 0.06%.
The diff coverage is 85.71%.

@@            Coverage Diff             @@
##           master     #752      +/-   ##
==========================================
- Coverage   83.52%   83.45%   -0.07%     
==========================================
  Files          72       72              
  Lines        5541     5568      +27     
==========================================
+ Hits         4628     4647      +19     
- Misses        913      921       +8     

Impacted Files	Coverage Δ
lstchain/io/event_selection.py	100.00% <ø> (ø)
lstchain/reco/utils.py	70.45% <44.44%> (-2.97%)	⬇️
lstchain/irf/hdu_table.py	92.18% <95.55%> (+0.64%)	⬆️
lstchain/irf/__init__.py	100.00% <100.00%> (ø)
lstchain/irf/tests/test_hdus.py	100.00% <100.00%> (ø)
lstchain/reco/tests/test_utils.py	100.00% <100.00%> (ø)
lstchain/tools/lstchain_create_dl3_file.py	90.62% <100.00%> (+0.14%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f49becb...e67d553. Read the comment docs.

[moralejo]

Looks ok to me - assuming it has been tested on the usual reference data samples