Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-17597

Update testdata_jointcal catalogs to be processed with gaia+ps1

    Details

    • Type: Story
    • Status: Done
    • Resolution: Done
    • Fix Version/s: None
    • Component/s: jointcal
    • Labels:
      None
    • Story Points:
      28
    • Sprint:
      AP S20-5 (April), AP S20-6 (May)
    • Team:
      Alert Production
    • Urgent?:
      No

      Description

      At some point, we should update the jointcal testdata to reflect the latest processCcd output (e.g. using the new Pessimistic matcher) processed with a more modern reference catalog (likely PS1 for photometry, Gaia for astrometry). This should allow us to un-skip several of the existing tests (e.g. DeCAM photometry tests) that are currently skipped because of bad test data, and remove old-style sourceSelector flags, etc. This would also be the time to update testdata_jointcal to the gen3 butler format.

      As part of this work, we should investigate in detail the various chi2/ndof values computed by the tests (these were changed as part of DM-17595, without in-depth study). This would be the time to really test assumptions about e.g. no_rank_update or line search astrometry tests not changing the default chi2, when a highly accurate reference catalog is used (Gaia).

      I would expect this work to be delayed for a year or so, until all of the above pieces are nicely settled into place.

        Attachments

          Issue Links

            Activity

            Hide
            erykoff Eli Rykoff added a comment -

            Well, even if the tests values don't change then the configs will have to be changed to reflect the new data formats (overrides can be removed, yay!).

            Show
            erykoff Eli Rykoff added a comment - Well, even if the tests values don't change then the configs will have to be changed to reflect the new data formats (overrides can be removed, yay!).
            Show
            Parejkoj John Parejko added a comment - - edited Jenkins run: https://ci.lsst.codes/blue/organizations/jenkins/stack-os-matrix/detail/stack-os-matrix/31934/pipeline
            Hide
            Parejkoj John Parejko added a comment - - edited

            Eli Rykoff are you still ok to review this? the testdata claims it's a tens of thousands of lines, but that's because of the data (please do have a look at the compression scripts and the README changes though!). The biggest change is jointcal itself, where in addition to the test value changes, I had to restructure some calculations. The commits in general should be atomic, and reviewing by commit may make it all more manageable.

            Here are the relevant PRs:

            Show
            Parejkoj John Parejko added a comment - - edited Eli Rykoff are you still ok to review this? the testdata claims it's a tens of thousands of lines, but that's because of the data (please do have a look at the compression scripts and the README changes though!). The biggest change is jointcal itself, where in addition to the test value changes, I had to restructure some calculations. The commits in general should be atomic, and reviewing by commit may make it all more manageable. Here are the relevant PRs: https://github.com/lsst/testdata_jointcal/pull/26 https://github.com/lsst/jointcal/pull/157 https://github.com/lsst/obs_decam/pull/153 https://github.com/lsst/validate_drp/pull/122 https://github.com/lsst/fgcmcal/pull/37 (I'll review this one, since you wrote it)
            Hide
            erykoff Eli Rykoff added a comment -

            My part of the review is complete, but we need somebody (you, John Parejko?) to look at the fgcmcal PR.

            Show
            erykoff Eli Rykoff added a comment - My part of the review is complete, but we need somebody (you, John Parejko ?) to look at the fgcmcal PR.
            Hide
            Parejkoj John Parejko added a comment -

            Thanks for the review, Eli Rykoff. I merged everything except obs_cfht, which I just closed with a note about RFC-697.

            Merged and done, after far too long!

            Show
            Parejkoj John Parejko added a comment - Thanks for the review, Eli Rykoff . I merged everything except obs_cfht, which I just closed with a note about RFC-697 . Merged and done, after far too long!

              People

              • Assignee:
                Parejkoj John Parejko
                Reporter:
                Parejkoj John Parejko
                Reviewers:
                Eli Rykoff
                Watchers:
                Eli Rykoff, Jim Bosch, John Parejko, John Swinbank
              • Votes:
                0 Vote for this issue
                Watchers:
                4 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved:

                  Summary Panel