I’m currently using the object catalog in EDP2, particularly the light curves in objectForcedSource. I’m a bit confused because some objects have ~40 epochs in their light curves, with ~30 filled psfMag values, but most of the points have an empty/NaN psfMagErr. For example, objectId: 756398235370674612 has 44 epochs in its light curve, 24 of the psfMag values are null, and 40 of the psfMagErr values are null. None of these observations are flagged, so I was wondering what this means — are these observations still reliable? Thank you!
Hi @scharney, thanks for the question. I was unable to reproduce the issue. In the FourcedSource DP2 catalog for objectId= 756398235370674612, I see 46 measurements with no null entries for psfFlux or psfFluxErr using the Portal ADQL query:
SELECT objectId,visit,detector,band,psfFlux,psfFluxErr,psfDiffFlux,psfDiffFluxErr
FROM dp2.ForcedSource
WHERE (objectId =756398235370674612).
To add to Fritz’s diagnosis: the psfMag column is derived from psfFlux via psfMag = -2.5*log10(psfFlux) + zp, so any measurement with negative psfFlux comes out as null. The DP2 tutorial notebook 305.1 Variable star light curves is a good worked example.
For the source with objectId = 756398235370674612, I get the same results as @sgreenstreet showed above. Additionally, the ForcedSource table does not contain converted magnitude columns. So I guess you did the flux to magnitude conversion by yourself. In that case, @Garreth’s response seems to answer your question, so I have marked it as the solution. However, if this doesn’t fully resolve your issue, please feel free to unmark it and provide a bit more detail so we can reproduce the problem and assist you further.
