Meredith Rawls can you take a look at the coadds at /project/ebellm/coadd_2014/deepCoadd/g/0 and see if they appear correct? (I am on a very slow connection.)

There are... a lot of files in there. Do you know which ones correspond to which DECam field? I was kind of hoping for three FITS files that each had 60-however-many HDUs (one per DECam CCD), but that's probably not how coadds work.

I can pick Yusra's brain today or tomorrow since she is visiting and knows a lot about coadds. But any pointers to which files or subdirectories you think have the useful information would be helpful too!

After working with Yusra a bit, I've gotten a non-PSF-matched coadd (called a "direct coadd") to work as a template with image differencing!

However, it would be helpful if you could please give some more information on this ticket, such as...

• The exact arguments you used with makeCoaddTempExp.py and assembleCoadd.py to make your coadds
• The visits you used from the 2014 HiTS dataset (for each field) to go into the coadds and how you chose them

According to Yusra AlSayyad, we can reduce the size of the final coadds (the ones that will actually go into ap_verify_hits2015) by about 90% if we delete the warps from the coadd output repository. This should lead to a more manageable package size.

Superseded by DM-15080.