Validating `full-enclosure` DL3 files

TarekHC commented 4 years ago

As in #18 , @mireianievas and I want to know what needs to be done to validate full-enclosure IRFs. This validation is more tricky, because 3D analysis is not standard at all. Both gammapy and ctools may be used (although the manpower is not infinite... So we might stick with one). Probably we will need IRFs generated for all offsets (probably using GRISU).

[ ] Make sure we are able to reconstruct the Crab spectrum
[ ] Compare skymaps (e.g. significance distributions).
[ ] Compare significance vs time.
[ ] a steep blazar (e.g. PKS 1414)
[ ] an empty field (e.g. Segue I or other dSphs)
[ ] a week source (e.g. Tycho)

Any other test?

mireianievas commented 4 years ago

I'm generating a bunch of EA for GRISU right now. They should cover all the V4 - V6 2019 (including the size-scale related IRFs for V6) and the different offsets.

GernotMaier commented 4 years ago

I would suggest to add three more sources types:

a steep blazar (e.g. PKS 1414; quite strong, so not many runs are needed)
an empty field
a week source (like Tycho; do this last, as you will many runs for it)

GernotMaier commented 3 years ago

Close this - needs to be done on the V2DL3 / gammapy discussion level.

VERITAS-Observatory / EventDisplay_v4

Validating `full-enclosure` DL3 files #19