Let's try to capture what are the issues/concerns about the data before resuming 2.1i:
[ ] PSF size (issue #313 )
[x] PSF residual correlation: In a single visit, when compared to the LSST-SRD we don't pass the TEx criteria. Caveat: The criteria is for full survey data. After discussing with @rmjarvis, he pointed out that the values should decrease as ~N (N number of exposures) in the coadds. If we increase the tolerance by a factor of ~100, the measured residuals pass the criteria.
Absolute astrometry and photometry are in compliance with the SRD.
Depth in compliance with the SRD and OpSim expectation. Concern: We'll have to re-check this when the new PSFs are in place since a larger PSF should reduce the depth.
PSF ellipticity, |e|, in compliance with the maximum values in the SRD. When we change the PSF we will have to re-check this and see if the distribution still looks good.
All, are there other issues/concerns? What other tests can we do with these calexps? Are there other checks to be done using the centroid files?
To do:
[ ] Try to run validate_drp to cross-check the results on the TEx criteria.
Let's try to capture what are the issues/concerns about the data before resuming 2.1i:
[x] PSF residual correlation: In a single visit, when compared to the LSST-SRD we don't pass the TEx criteria. Caveat: The criteria is for full survey data. After discussing with @rmjarvis, he pointed out that the values should decrease as ~N (N number of exposures) in the coadds. If we increase the tolerance by a factor of ~100, the measured residuals pass the criteria.
Absolute astrometry and photometry are in compliance with the SRD.
Depth in compliance with the SRD and OpSim expectation. Concern: We'll have to re-check this when the new PSFs are in place since a larger PSF should reduce the depth.
PSF ellipticity, |e|, in compliance with the maximum values in the SRD. When we change the PSF we will have to re-check this and see if the distribution still looks good.
All, are there other issues/concerns? What other tests can we do with these
calexps
? Are there other checks to be done using thecentroid
files?To do:
validate_drp
to cross-check the results on the TEx criteria.