As part of this, I'm also pulling in 78685e8 from HSC's meas_algorithms, which does not have an HSC-side ticket.

Also 88196ac and d90603d from HSC.

Also 5c44605. This was part of HSC-551; the remainder of that ticket has already been merged to meas_multifit as part of DM-240.

Hey Lauren, do you have time for a review, please?

The changes here are fairly short, but they do span quite a few repositories. In particular, please see the tickets/DM-3136 branch on:

• afw
• coadd_utils
• meas_base
• meas_algorithms
• ip_diffim
• pipe_tasks
• obs_subaru
• lsst_dm_stack_demo

Note that I can't run the changes to the stack demo through Buildbot before merging, as it only pulls the master branch. I have tested them locally, though. Similarly, Buildbot does nothing at all with obs_subaru.

All looks good to me. Did you check the new demo outputs against the old ones to make sure the former make sense (I imagine the changes should be minor for the most part, but perhaps at least a few cases will change significantly with the new threshold computation/masking?)

Otherwise, modulo a few minor comments in the commits (they weren't inter-line, so I didn't bother with a pull request), you are free to merge.