The current implementation of SkyModel.get_MIGHTEE_Sky has some major errors from our side, because we don't extract and convert the sources and their parameters correctly from the .fits file. Our SkyModel treats the sources as follows:
We do the following not correctly:
Assignment of Stokes-I flux, currently, "NU_EFF" is taken, which are actually frequencies in HZ and not peak-Flux in Jy/beam (stored in "S_PEAK")
major & minor axis of SkyModel are in arcsec, however, if the unit is different, like it is for the MIGHTEE sky (in deg), we need to convert them.
Frequencies are just taken from SkyModel.get_sky_model_from_fits as input parameters and not from the .fits file. But they're availabe at "NU_EFF" and we just ignore them. They should be passed to reference_frequency.
As far as I understand, these are the errors we do for the MIGHTEE sky. @rohitcbscient please correct me if I'm mistaken in my statements or referencing the correct header-names.
As a hotfix, we should just raise an error, until this is fixed. Otherwise, you just get a corrupt sky.
The current implementation of
SkyModel.get_MIGHTEE_Skyhas some major errors from our side, because we don't extract and convert the sources and their parameters correctly from the .fits file. OurSkyModeltreats the sources as follows:We do the following not correctly:
SkyModelare in arcsec, however, if the unit is different, like it is for the MIGHTEE sky (in deg), we need to convert them.SkyModel.get_sky_model_from_fitsas input parameters and not from the .fits file. But they're availabe at "NU_EFF" and we just ignore them. They should be passed toreference_frequency.As far as I understand, these are the errors we do for the MIGHTEE sky. @rohitcbscient please correct me if I'm mistaken in my statements or referencing the correct header-names.
As a hotfix, we should just raise an error, until this is fixed. Otherwise, you just get a corrupt sky.