Closed Will-Cooper closed 2 years ago
Similarly, checking the flux unit is a correct measure (e.g. can be converted correctly like erg/ cm^2/ A/ s)
I have a fix for the flux units in place, but looking at that object I don't see pixel wavelengths:
data = db.inventory('2MASS J00325584-4405058')
for x in data['Spectra']:
print(x['instrument'], x['wavelength_units'], x['flux_units'], x['spectrum'])
SpeX um ergs-1cm-2A-1 https://s3.amazonaws.com/bdnyc/SpeX/Prism/spex_prism_0032-4405_U20026.fits
ISAAC None None https://bdnyc.s3.amazonaws.com/ISAAC/2M0032_ISAACspec_JHK_nocosmics.fits
R-C Spec A ergs-1cm-2A-1 https://s3.amazonaws.com/bdnyc/optical_spectra/U20026.fits
LRIS A ergs-1cm-2A-1 https://s3.amazonaws.com/bdnyc/optical_spectra/U20026_0032-4405.fits
The ISSAC spectrum can't be read by specutils at all due to the format so that is the main issue.
We need someway to check when ingesting spectra, that the spectrum will be able to be loading when being parsed downstream. For example 2MASS J00325584-4405058 has wavelength units in pixels, which can't really be plotted (alongside anything else at least) on the website. Perhaps in the
ingest_utils
code when looping through the objects,try
to load the object into astrodbkit, then check the unit is some measure of distance unit.