LSSTDESC / descwl-shear-sims

simple simulations for testing weak lensing shear measurement
BSD 3-Clause "New" or "Revised" License
10 stars 9 forks source link

make sim return exposures #84

Closed esheldon closed 3 years ago

esheldon commented 4 years ago

That is what we will get off disk when we convert to butler, better to make the sim deliver that.

Also, that is the only way we will be able to keep straight the mask bits.

beckermr commented 4 years ago

Not following here. The sims have some metadata associated with them that may not be amenable to an LSST exposure, right?

esheldon commented 4 years ago

We should return additional metadata, but use an exposure to store the image, bmask, variance plane.

esheldon commented 3 years ago

This has bumped up to high priority. It is important that the downstream code in descwl_coadd takes in exposures so that we don't have different code paths for simluated and real data

esheldon commented 3 years ago

For variable psfs, we are evaluating it at the center of the single epoch image and using that in a FixedPSF. We had intended to evaluate it at the coadd center, so this seems like a bug. I will try to fix, but I guess this means we are not very sensitive to this.

To make the code path consistent with what we'll do with real data, we need to return a mocked up DM psf and make sure to evaluate at the coadd center in the coadding code

esheldon commented 3 years ago

Note it looks like we are doing the right thing in the coadd code.

beckermr commented 3 years ago

For your previous comment where is that code snippet? I think this must be the rewritten code and not what I made originally.

esheldon commented 3 years ago

no doubt this is my mistake: https://github.com/LSSTDESC/descwl-shear-sims/blob/master/descwl_shear_sims/sim/objlists.py#L87

esheldon commented 3 years ago

We need to also return a bbox rather than coadd sims, see https://github.com/LSSTDESC/descwl_coadd/issues/45