Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-8111

Exercise prototype on astrometry failure modes

    XMLWordPrintable

    Details

    • Story Points:
      6
    • Sprint:
      Alert Production F16 - 11, Alert Production F16 - 11b, Alert Production F16 - 11c, Alert Production S17 - 12
    • Team:
      Alert Production

      Description

      Run the python implementation of OPMb on data with known failure modes with the previous matcher.

      Deliverables: Pass/Fail for new code.

        Attachments

          Activity

          Hide
          cmorrison Chris Morrison [X] (Inactive) added a comment -

          Started work on real data examples with both the current implementation of the pattern matcher and the pure python matcher.

          Show
          cmorrison Chris Morrison [X] (Inactive) added a comment - Started work on real data examples with both the current implementation of the pattern matcher and the pure python matcher.
          Hide
          cmorrison Chris Morrison [X] (Inactive) added a comment -

          Exercised the 3D, pure python match optimistic b implementation on CFHT visits: 793169, 980526, 89070; DECam visit: 303497; and HSC visit 3358. The CFHT and HSC visits are known failure modes from DM-8109 and DM-8110.

          CFHT: The new matcher performance is slightly mixed compared to the old matcher due to the high distortions and low reference object count(especially in the u band visits). This is likely due to the the high dynamic ratio the current matcher need to fit both CFHT, DECam, and the HSC data.

          DECam: The new matcher produces wcs fits of similar quality to the old matcher.

          HSC: This python version of the matcher succeeds on 42 out of the 112 CCDs in the HSC visit where as the old matcher succeeded on only a handful of the frames. "success" here is defined as a final RMS scatter on the matchers of <0.2 arcseconds. The new matcher also has some intermediate successes where the matcher RMS is <1.0 and would like succeed if the match_and_fit_wcs iterations were more. Current default is 3.

          Both matchers appear to be very sensitive to the start points of the match search(max shift allowed, rotation angle, match distance, etc.).

          Show
          cmorrison Chris Morrison [X] (Inactive) added a comment - Exercised the 3D, pure python match optimistic b implementation on CFHT visits: 793169, 980526, 89070; DECam visit: 303497; and HSC visit 3358. The CFHT and HSC visits are known failure modes from DM-8109 and DM-8110 . CFHT: The new matcher performance is slightly mixed compared to the old matcher due to the high distortions and low reference object count(especially in the u band visits). This is likely due to the the high dynamic ratio the current matcher need to fit both CFHT, DECam, and the HSC data. DECam: The new matcher produces wcs fits of similar quality to the old matcher. HSC: This python version of the matcher succeeds on 42 out of the 112 CCDs in the HSC visit where as the old matcher succeeded on only a handful of the frames. "success" here is defined as a final RMS scatter on the matchers of <0.2 arcseconds. The new matcher also has some intermediate successes where the matcher RMS is <1.0 and would like succeed if the match_and_fit_wcs iterations were more. Current default is 3. Both matchers appear to be very sensitive to the start points of the match search(max shift allowed, rotation angle, match distance, etc.).
          Hide
          ctslater Colin Slater added a comment -

          There's not much for me to review about this; looks good to me.

          In retrospect, I think the ideal workflow here is probably to have a DMTN up and running, and then this ticket could commit the results to that text. But as is, I think it's fine to leave this here with the intention of the results making their way into the DMTN later.

          Show
          ctslater Colin Slater added a comment - There's not much for me to review about this; looks good to me. In retrospect, I think the ideal workflow here is probably to have a DMTN up and running, and then this ticket could commit the results to that text. But as is, I think it's fine to leave this here with the intention of the results making their way into the DMTN later.
          Hide
          cmorrison Chris Morrison [X] (Inactive) added a comment -

          Finished processing and review. Next need to sumarize these results into the DMTNs from DM-8114.

          Show
          cmorrison Chris Morrison [X] (Inactive) added a comment - Finished processing and review. Next need to sumarize these results into the DMTNs from DM-8114 .

            People

            Assignee:
            cmorrison Chris Morrison [X] (Inactive)
            Reporter:
            cmorrison Chris Morrison [X] (Inactive)
            Reviewers:
            Colin Slater, Simon Krughoff
            Watchers:
            Chris Morrison [X] (Inactive), Colin Slater, Simon Krughoff
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved:

                Jenkins Builds

                No builds found.