Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-21702

Investigate the cause of the high number of false sources in convolveTemplate=False image differencing

    XMLWordPrintable

    Details

    • Story Points:
      0
    • Epic Link:
    • Sprint:
      AP F19-5 (October), AP F19-6 (November)
    • Team:
      Alert Production

      Description

      ip_diffim has the bool configuration option convolveTemplate implemented. With convolveTemplate=False, it switches the roles of the science and template images and performs the AL convolution of the science image to match the template. For cases when the science image is sharper than the template convolveTemplate=False and doDecorrelation=True setup should perform as well as the "baseline" case when the template is the sharper one and it is convolved.

       

      The resulting difference image is bad quality, indicated by an order of magnitude (~ 10 times) more sources detected (half positive, half negative approximately).

      This ticket covers the work by Gabor Kovacs [X] to investigate the cause of this bug.

        Attachments

          Issue Links

          is parent task of

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21703 Run the science image convolution image differencing for a visit when it is not a close matching PSF case

          • Done

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21704 Check whether the template scaling has a connection to the number of detected souorces

          • Done

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21705 Use the correct kernelSize in case of science image convolution in image differencing

          • Done

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21809 Check the variance plane operations and the decorrelation afterburner operations in the science convolution case

          • Done
          relates to

          Bug - A problem which impairs or prevents the functions of the product. DM-21700 Photometric scaling of fluxes differ between template coadds and calexps

          • Won't Fix

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21017 Consider changing the automated algorithmic path in ip_diffim to disable automated deconvolution

          • To Do

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21217 Add option to select image to convolve based on seeing

          • To Do

          Story - Created by Jira Software - do not edit or delete. Issue type for a user story. DM-21920 Try out and write notebook about demonstrating basic DFT space steps for ip_diffim operations

          • Done

            Activity

            Ascending order - Click to sort in descending order
            No builds found.
            gkovacs Gabor Kovacs [X] (Inactive) created issue -
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Field Original Value New Value
            Link This issue relates to DM-21017 [ DM-21017 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue relates to DM-21217 [ DM-21217 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Summary Investigate the cause of the high number of detected sources in the convolveScience=True image differencing case Investigate the cause of the high number of false sources in convolveTemplate=False image differencing
            Hide
            Permalink
            gkovacs Gabor Kovacs [X] (Inactive) added a comment - - edited

            DM-21700 Theoretically, this should not affect the AL subtraction.

            Show
            gkovacs Gabor Kovacs [X] (Inactive) added a comment - - edited DM-21700 Theoretically, this should not affect the AL subtraction.
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue relates to DM-21700 [ DM-21700 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue is parent task of DM-21703 [ DM-21703 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue is parent task of DM-21704 [ DM-21704 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue is parent task of DM-21705 [ DM-21705 ]
            swinbank John Swinbank made changes -
            Description {{ip_diffim}} has the bool configuration option {{convolveTemplate}} implemented. With {{convolveTemplate=False}}, it switches the roles of the science and template images and performs the AL convolution of _the science image_ to match _the template_. For cases when the science image is sharper than the template {{convolveTemplate=False}} and {{doDecorrelation=True}} setup should perform as good as the "baseline" case when the template is the sharper one and it is convolved.

             

            The resulting difference image is bad quality, indicated by an order of magnitude (~ 10 times) more sources detected (half positive, half negative approximately).

            This ticket covers the work by [~gkovacs] to investigate the cause of this bug.             		{{ip_diffim}} has the bool configuration option {{convolveTemplate}} implemented. With {{convolveTemplate=False}}, it switches the roles of the science and template images and performs the AL convolution of _the science image_ to match _the template_. For cases when the science image is sharper than the template {{convolveTemplate=False}} and {{doDecorrelation=True}} setup should perform as well as the "baseline" case when the template is the sharper one and it is convolved.

             

            The resulting difference image is bad quality, indicated by an order of magnitude (~ 10 times) more sources detected (half positive, half negative approximately).

            This ticket covers the work by [~gkovacs] to investigate the cause of this bug.
            swinbank John Swinbank made changes -
            Epic Link DM-21444 [ 423051 ]
            swinbank John Swinbank made changes -
            Team Alert Production [ 10300 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue is parent task of DM-21809 [ DM-21809 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Status To Do [ 10001 ] In Progress [ 3 ]
            Hide
            Permalink
            gkovacs Gabor Kovacs [X] (Inactive) added a comment -

            This is just the cover ticket of my experiments with the science image convolution. Can we close it perhaps at the sprint planning?

            Show
            gkovacs Gabor Kovacs [X] (Inactive) added a comment - This is just the cover ticket of my experiments with the science image convolution. Can we close it perhaps at the sprint planning?
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Reviewers Eric Bellm [ ebellm ]
            Status In Progress [ 3 ] In Review [ 10004 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Link This issue relates to DM-21920 [ DM-21920 ]
            ebellm Eric Bellm made changes -
            Sprint AP F19-5 (October) [ 957 ]
            Story Points 2
            ebellm Eric Bellm made changes -
            Story Points 2 0
            ebellm Eric Bellm made changes -
            Component/s ip_diffim [ 10743 ]
            Labels SciencePipelines
            ebellm Eric Bellm made changes -
            Status In Review [ 10004 ] Reviewed [ 10101 ]
            ebellm Eric Bellm made changes -
            Sprint AP F19-5 (October) [ 957 ] AP F19-5 (October), AP F19-6 (November) [ 957, 958 ]
            gkovacs Gabor Kovacs [X] (Inactive) made changes -
            Resolution Done [ 10000 ]
            Status Reviewed [ 10101 ] Done [ 10002 ]

              People

              Assignee:
              gkovacs Gabor Kovacs [X] (Inactive)
              Reporter:
              gkovacs Gabor Kovacs [X] (Inactive)
              Reviewers:
              Eric Bellm
              Watchers:
              Eric Bellm, Gabor Kovacs [X] (Inactive)
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                  Jenkins

                  No builds found.