mdebony / acceptance_modelisation Goto Github PK

I really like the idea of using the FoV rotation to calculate the time frame in which to evaluate the observations 😊
Unfortunately I came across the problem that this can fail in the case that the time frame is so small that there are no events in the sub-observation you are looking at.
I guess one solution would be to check the number events for the time frame but this requires to do obs.select_time() inside the _compute_time_intervals_based_on_fov_rotation method. The other solution would be to required something like a deltaT_threshold so that there are no frames that have smaller duration than this threshold :)

Astropy has the ErfaAstromInterpolator, which makes it possbile to interpolate some of the transformation related things. Otherwise it needs to to these for every event as they all have individual timestamps.
See:
https://docs.astropy.org/en/latest/coordinates/index.html

Hi 😊

I think the fov_alignment has to be given at the instantiation of Background3D here, e.g. fov_alignment=FoVAlignment.ALTAZ

Otherwise it's RADEC per default and as far as I can tell by roughly going through the code, the acceptance model is calculated in the SkyOffset frame of the camera in altaz