Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-4063

Support new casting requirements in NumPy 1.10

    Details

    • Story Points:
      1
    • Epic Link:
    • Sprint:
      Science Pipelines DM-W16-2
    • Team:
      Alert Production

      Description

      The function imagesDiffer() in testUtils attempts to OR an array of unit16s (LHS) against an array of bools(RHS) valSkipMaskArr |= skipMaskArr and errors with message

      TypeError: ufunc 'bitwise_or' output (typecode 'H') could not be coerced to provided output parameter (typecode '?') according to the casting rule ''same_kind''
      

      preventing afw from building correctly.

        Attachments

          Issue Links

            Activity

            Hide
            rowen Russell Owen added a comment -

            I merged afw and coadd_utils, but we are also seeing the same problem in obs_lsstSim tests/testLsstSimIsrTask.py:

              File "/Users/rowen/LSST/lsstsw/build/obs_lsstSim/python/lsst/obs/lsstSim/lsstSimIsrTask.py", line 81, in saturationInterpolation
                self.unmaskSatHotPixels(ccdExposure)
              File "/Users/rowen/LSST/lsstsw/build/obs_lsstSim/python/lsst/obs/lsstSim/lsstSimIsrTask.py", line 70, in unmaskSatHotPixels
                maskarr[idx] &= andMask
            TypeError: ufunc 'bitwise_and' output (typecode 'i') could not be coerced to provided output parameter (typecode 'H') according to the casting rule ''same_kind''
            

            Show
            rowen Russell Owen added a comment - I merged afw and coadd_utils, but we are also seeing the same problem in obs_lsstSim tests/testLsstSimIsrTask.py: File "/Users/rowen/LSST/lsstsw/build/obs_lsstSim/python/lsst/obs/lsstSim/lsstSimIsrTask.py", line 81, in saturationInterpolation self.unmaskSatHotPixels(ccdExposure) File "/Users/rowen/LSST/lsstsw/build/obs_lsstSim/python/lsst/obs/lsstSim/lsstSimIsrTask.py", line 70, in unmaskSatHotPixels maskarr[idx] &= andMask TypeError: ufunc 'bitwise_and' output (typecode 'i') could not be coerced to provided output parameter (typecode 'H') according to the casting rule ''same_kind''
            Hide
            rowen Russell Owen added a comment -

            Could you please also look at obs_lsstSim tickets/DM-4063? A change much like the others.

            At this point all of lsst_apps builds for me. I tried building lsst_sims but a package hung for no obvious reason (though I would be surprised if it had anything to do with the numpy change) and I don't have time to diagnose it yet.

            Show
            rowen Russell Owen added a comment - Could you please also look at obs_lsstSim tickets/ DM-4063 ? A change much like the others. At this point all of lsst_apps builds for me. I tried building lsst_sims but a package hung for no obvious reason (though I would be surprised if it had anything to do with the numpy change) and I don't have time to diagnose it yet.
            Hide
            tjenness Tim Jenness added a comment -

            The obs_lsstSim change is like the others so seems fine for now. Thanks.

            Show
            tjenness Tim Jenness added a comment - The obs_lsstSim change is like the others so seems fine for now. Thanks.
            Hide
            rowen Russell Owen added a comment -

            I merged obs_lsstSim. At this point all of lsst_apps builds for on my Mac using numpy 1.10 and Peter Yoachim has kindly agreed to take up any issues in sims packages using a different ticket.

            Show
            rowen Russell Owen added a comment - I merged obs_lsstSim. At this point all of lsst_apps builds for on my Mac using numpy 1.10 and Peter Yoachim has kindly agreed to take up any issues in sims packages using a different ticket.
            Hide
            vpk24 Vishal Kasliwal [X] (Inactive) added a comment -

            I can confirm that all of lsst_apps builds on ubuntu.

            Show
            vpk24 Vishal Kasliwal [X] (Inactive) added a comment - I can confirm that all of lsst_apps builds on ubuntu.

              People

              • Assignee:
                rowen Russell Owen
                Reporter:
                vpk24 Vishal Kasliwal [X] (Inactive)
                Reviewers:
                Tim Jenness
                Watchers:
                John Swinbank, Russell Owen, Tim Jenness, Vishal Kasliwal [X] (Inactive)
              • Votes:
                0 Vote for this issue
                Watchers:
                4 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: