Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-20154

Implement new initial WCS design

    XMLWordPrintable

Details

    • 8
    • AP S19-6, AP F19-1, AP F19-2
    • Alert Production

    Description

      The close of DM-19951 describes a new design for how we create an initial WCS from the raw exposure VisitInfo metadata. This ticket is to implement that design. In summary:

      • Write tests against sensor bounding boxes on known data (e.g. testdata_*).
      • Write a new function to create a SkyWcs from VisitInfo and Detector (createInitialSkyWcs?).
      • Call that function from exposureFromImage.
      • Reorder the reading of raw VisitInfo and WCS metadata to deal potential metadata stripping problems.
      • Check that all existing cameras use this code path, and not some custom Exposure metadata handling.
      • Turn off addDistortionModel in ISR for all cameras, so that the above WCS does not get overwritten.

      Attachments

        Issue Links

          Activity

            After lots of fun discussions about docstring inheritance, CRVAL being the actual worst, deprecation warnings, etc., I'm marking this as reviewed pending an ap_verify run on the CI dataset demonstrating the metrics that do not wildly differ from a recent ap_verify actual-CI run. Let me know if you want help with that!

            mrawls Meredith Rawls added a comment - After lots of fun discussions about docstring inheritance, CRVAL being the actual worst, deprecation warnings, etc., I'm marking this as reviewed pending an ap_verify run on the CI dataset demonstrating the metrics that do not wildly differ from a recent ap_verify actual-CI run. Let me know if you want help with that!
            Parejkoj John Parejko added a comment - - edited

            I ran ap_verify on ap_verify_ci_hits2015, and the output seems to match that of past runs (looking at the dashboard and one run against master, I get 1706 unassociated sources), so I think we're ok there.

            I'm still blocked on DM-20548, but otherwise I think we're ok to merge.

            Parejkoj John Parejko added a comment - - edited I ran ap_verify on ap_verify_ci_hits2015 , and the output seems to match that of past runs (looking at the dashboard and one run against master, I get 1706 unassociated sources), so I think we're ok there. I'm still blocked on DM-20548 , but otherwise I think we're ok to merge.
            Parejkoj John Parejko added a comment - Post-rebase Jenkins run: https://ci.lsst.codes/blue/organizations/jenkins/stack-os-matrix/detail/stack-os-matrix/30312/pipeline

            After rebasing, ap_verify_ci_hits2015 still produces the same metric, so we're still ok there. DM-20548 is merged, so once the above Jenkins run completes successfully, I'll finally merge this.

            Parejkoj John Parejko added a comment - After rebasing, ap_verify_ci_hits2015 still produces the same metric, so we're still ok there. DM-20548 is merged, so once the above Jenkins run completes successfully, I'll finally merge this.
            Parejkoj John Parejko added a comment -

            Thanks for all the review comments and testing suggestions mrawls.

            Merged and done.

            Parejkoj John Parejko added a comment - Thanks for all the review comments and testing suggestions mrawls . Merged and done.

            People

              Parejkoj John Parejko
              Parejkoj John Parejko
              Meredith Rawls
              Chris Morrison, Jim Bosch, John Parejko, John Swinbank, Lauren MacArthur, Meredith Rawls
              Votes:
              0 Vote for this issue
              Watchers:
              6 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Jenkins

                  No builds found.