How's the DIAObject association going?

Data Management
#1

Hi all! @rbiswas and I are reading up on the DM algorithms for variable object characterization, as part of the DESC Twinkles project. We’re interested to see how the DIAObject association pipeline (WBS 02C.03.02, page 32 of the LDM-151 algorithms document) will respond to pairs of supernovae, galaxies with both AGN and supernovae in them, and lensed variable sources (either multiply-imaged AGN or SNe). Does anyone have any brief news on this pipeline to share from the HSC analysis? Is there anything more up-to-date we can read than LDM-151? Thanks!

#2

@drphilmarshall we haven’t fleshed out the planning on this significantly. Currently we have a milestone that has this being completed in Summer of 2017 (see this link. Round boxes are milestones, square boxes represent the work necessary to produce the milestones).

Filling out the plan in LDM-151 (which may result in more milestones) is a necessity and input like this is very useful for doing that.

#3

Thanks Simon. That link is broken - looks like a local file. Are you saying that the documentation won’t be done until Summer 2017, or the DIAObject association pipeline won’t be implemented until Summer of 2017?

#4

By coincidence the machine hosting the link is down at the moment. It should be up in a bit.[quote=“drphilmarshall, post:3, topic:419”]
Are you saying that the documentation won’t be done until Summer 2017, or the DIAObject association pipeline won’t be implemented until Summer of 2017?
[/quote]
We are currently working on fleshing out LDM-151, so it will be done sooner. The association pipeline is slated to be done in Summer 2017, but if there are earlier milestones needed, we can add them.

#5

Some people working with HSC data have used the LSST ip_diffim code (version as of a year or two ago) to do image subtraction, and report that it appears to be working fine. It is not, however, currently run routinely as part of the official HSC Survey pipeline (we’ve been concentrating on wallpaper science so far).

#6

It’s up and running now. I can’t guarantee future uptime, but do please feel free to ping me directly if you spot it being down again.