Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-33821

Develop Chronograph metrics for known SSO attribution

    XMLWordPrintable

Details

    • Story
    • Status: Done
    • Resolution: Done
    • None
    • ap_verify

    Description

      After DM-32261, we should have solar system attribution running in CI. This ticket is to develop reasonable metrics for assessment and push them to Chronograph.

      Attachments

        Issue Links

          Activity

            ebellm, it looks like Chris Morrison wrote the following metrics:

            • numTotalSolarSystemObjects – the total number of SolarSystemObjects within a detectorVisit
            • numAssociatedSsObjects – number of SolarSystemObjects that were associated with new DiaSources for this detectorVisit

            Do we need any others?

            krzys Krzysztof Findeisen added a comment - ebellm , it looks like Chris Morrison wrote the following metrics: numTotalSolarSystemObjects – the total number of SolarSystemObjects within a detectorVisit numAssociatedSsObjects – number of SolarSystemObjects that were associated with new DiaSources for this detectorVisit Do we need any others?
            ebellm Eric Bellm added a comment -

            That seems like a reasonable start. Eventually it would be worthwhile to estimate how many known, detectable SSObjects were actually detected--but that will require more information about the intrinsic brightness of the SSObject that we don't have fully integrated now, I think.

            ebellm Eric Bellm added a comment - That seems like a reasonable start. Eventually it would be worthwhile to estimate how many known, detectable SSObjects were actually detected--but that will require more information about the intrinsic brightness of the SSObject that we don't have fully integrated now, I think.
            krzys Krzysztof Findeisen added a comment - - edited

            The metrics are now displayed on Chronograf, and have two interesting results:

            • There have been no successful associations since solar system data was added on March 1. This is true for both CI datasets.
            • The number of solar system objects registered as in-field (and therefore detectable) has fluctuated for HiTS twice, on March 30 [switch to Piff?] and May 4; those for COSMOS have been stable since March 1.
            krzys Krzysztof Findeisen added a comment - - edited The metrics are now displayed on Chronograf, and have two interesting results: There have been no successful associations since solar system data was added on March 1 . This is true for both CI datasets. The number of solar system objects registered as in-field (and therefore detectable) has fluctuated for HiTS twice, on March 30 [switch to Piff?] and May 4; those for COSMOS have been stable since March 1.

            Hi ebellm, can you take a look? In addition to the updates on SQuaSH, there's some docstring fixes in ap_association.

            krzys Krzysztof Findeisen added a comment - Hi ebellm , can you take a look? In addition to the updates on SQuaSH, there's some docstring fixes in ap_association .
            sullivan Ian Sullivan added a comment -

            The docstring updates in ap_association look fine. I did not find Squash-related code changes to review, but I can take another look if you point me to the right place.

            sullivan Ian Sullivan added a comment - The docstring updates in ap_association look fine. I did not find Squash-related code changes to review, but I can take another look if you point me to the right place.
            krzys Krzysztof Findeisen added a comment - - edited

            The catalog metrics dashboard has four new graphs below the fakes metrics. Note that "number of detectable objects" does vary over months-long timescales (different WCS processing?).

            krzys Krzysztof Findeisen added a comment - - edited The catalog metrics dashboard has four new graphs below the fakes metrics. Note that "number of detectable objects" does vary over months-long timescales (different WCS processing?).
            krzys Krzysztof Findeisen added a comment - - edited

            From sullivan on Slack:

            I had seen those plots before, and was happy with them. I think it's fine that they are per-CCD, since that will allow us to see in more fine-grained detail whether we are picking up the SSOs.

            krzys Krzysztof Findeisen added a comment - - edited From sullivan on Slack: I had seen those plots before, and was happy with them. I think it's fine that they are per-CCD, since that will allow us to see in more fine-grained detail whether we are picking up the SSOs.

            People

              krzys Krzysztof Findeisen
              ebellm Eric Bellm
              Ian Sullivan
              Eric Bellm, Ian Sullivan, Krzysztof Findeisen
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Jenkins

                  No builds found.