Erin Sheldon
Erin Sheldon
yes; there is no hurry for that, we won't need to run that code again until its needed for the deep field work
we should probably add a check that the psf pixel scale is not larger than the image pixel scale by more than a tiny amount (which can happen due to...
this can wait until after 2.0 comes out.
you should always add noise to the psf image when you will fit using LM (which the code is doing internally for `psf: fitgauss`. are you seeing psf fit failures?...
I did the main results of sheldon & huff with admom, so I'm a bit surprised.
oh, I thought you mean 0.001 bias with metacal
where do we stand on this one?
Here is an updated script from Anna Niemiec that shows the issue. Note changing the psf to 'fitgauss' removes the bias. The issue is somewhere in the `MetacalGaussPSF` code for...
Here is the current version of that. I probably introduced a bug recently, the original by @rmjarvis worked for distorted psfs https://github.com/esheldon/ngmix/blob/51c488b0914a4499b9372875576a463e2697b29d/ngmix/metacal/metacal.py#L481
If I do not reconvolve the inferred gaussian PSF by the pixel it works.