Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-20556

collect false positive sprint visualizations

    Details

    • Type: Story
    • Status: Done
    • Resolution: Done
    • Fix Version/s: None
    • Component/s: None
    • Story Points:
      12
    • Epic Link:
    • Sprint:
      AP F19-1, AP F19-4 (2H September), AP F19-5 (October), AP F19-6 (November), AP S20-1 (December), AP S20-2 (January), AP S20-3 (February), AP S20-4 (March)
    • Team:
      Alert Production

      Description

      Yusra AlSayyad produced some nice visualizations as part of the June False Positive print (e.g., https://github.com/lsst-dm/ap_pipe-notebooks/blob/master/false_positives/DiaSourceCensusCcdVisitNight.ipynb). This ticket is to extract them from notebooks into a more generalized script or library

        Attachments

          Issue Links

            Activity

            Hide
            mrawls Meredith Rawls added a comment -

            I added a bunch of plotting functions to ap_pipe-notebooks, and put examples of all the different plots one can make with just an APDB (well, and sometimes a butler) in a notebook.

            Show
            mrawls Meredith Rawls added a comment - I added a bunch of plotting functions to ap_pipe-notebooks, and put examples of all the different plots one can make with just an APDB (well, and sometimes a butler) in a notebook.
            Hide
            sophiereed Sophie Reed added a comment -

            A few code comments on github and some general plot appearance comments below.

            It would be nice if the colour maps differentiated between no possible data (outside the ccds) and no value here for other reasons. Maybe make it not start at white/very pale blue?

            It would be good if you could read the individual visit numbers so you can tell which ones correspond to outliers if necessary. Might be too hard with so many data points though.

            It would look nicer if the ccd numbers were centred and if they were outlined in white so they could be easily read against the dark blue.

            For some of the plots it would be worth making the titles/axis labels/tick labels bigger so that they can just be used straight off in a slide without having to make fake labels or endure the inevitable, "what's on the axis?".

             

            Show
            sophiereed Sophie Reed added a comment - A few code comments on github and some general plot appearance comments below. It would be nice if the colour maps differentiated between no possible data (outside the ccds) and no value here for other reasons. Maybe make it not start at white/very pale blue? It would be good if you could read the individual visit numbers so you can tell which ones correspond to outliers if necessary. Might be too hard with so many data points though. It would look nicer if the ccd numbers were centred and if they were outlined in white so they could be easily read against the dark blue. For some of the plots it would be worth making the titles/axis labels/tick labels bigger so that they can just be used straight off in a slide without having to make fake labels or endure the inevitable, "what's on the axis?".  
            Hide
            sophiereed Sophie Reed added a comment -

            When plotting the cutout I think it would be cleaner to have no axis ticks and labels unless the values mean something in which case it might be better to display them in an easier to read manner. A scale bar might be helpful but probably isn't necessary. 

            Show
            sophiereed Sophie Reed added a comment - When plotting the cutout I think it would be cleaner to have no axis ticks and labels unless the values mean something in which case it might be better to display them in an easier to read manner. A scale bar might be helpful but probably isn't necessary. 
            Hide
            mrawls Meredith Rawls added a comment -

            Thanks for the review. I addressed the code comments and also went ahead and made everything work with HSC. Along the way I discovered I had accidentally reversed "area per CCD" with "area per visit," whoops, but now I compute it instead of hardwiring it, so I suppose that's a win. I adjusted a couple of the text sizes but chose to leave the color scheme and other things as-is. I agree it's helpful to distinguish between no possible data and no sources but everything I tried made it look uglier. Do you want to take another look?

            Show
            mrawls Meredith Rawls added a comment - Thanks for the review. I addressed the code comments and also went ahead and made everything work with HSC. Along the way I discovered I had accidentally reversed "area per CCD" with "area per visit," whoops, but now I compute it instead of hardwiring it, so I suppose that's a win. I adjusted a couple of the text sizes but chose to leave the color scheme and other things as-is. I agree it's helpful to distinguish between no possible data and no sources but everything I tried made it look uglier. Do you want to take another look?
            Hide
            mrawls Meredith Rawls added a comment -

            Merged after 1:1 slack chat with Sophie.

            Show
            mrawls Meredith Rawls added a comment - Merged after 1:1 slack chat with Sophie.

              People

              • Assignee:
                mrawls Meredith Rawls
                Reporter:
                ebellm Eric Bellm
                Reviewers:
                Sophie Reed
                Watchers:
                Eric Bellm, Meredith Rawls, Sophie Reed
              • Votes:
                0 Vote for this issue
                Watchers:
                3 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved:

                  Summary Panel