Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-4344

pipe_tasks 11.0-14-ga314014+5 fails a test on os/x 10.10.5

    XMLWordPrintable

Details

    • Bug
    • Status: Done
    • Resolution: Done
    • None
    • pipe_tasks
    • None
    • 0
    • Science Pipelines DM-W16-4
    • Alert Production

    Description

      Running

      SCONSFLAGS="-j 6 opt=3" eups distrib install pipe_tasks 11.0-14-ga314014+5
      

      on my laptop running ox/x 10.10.5 results in a test failure:

      F.
      ======================================================================
      FAIL: testEdge (__main__.interpolationTestCase)
      Test that we can interpolate to the edge
      ----------------------------------------------------------------------
      Traceback (most recent call last):
        File "tests/testInterpImageTask.py", line 127, in testEdge
          validateInterp(miInterp, useFallbackValueAtEdge, fallbackValue)
        File "tests/testInterpImageTask.py", line 90, in validateInterp
          self.assertAlmostEqual(val0, fallbackValue, 6)
      AssertionError: 2.5515668 != 2.5515660972664267 within 6 places
      

      Attachments

        Issue Links

          Activity

            tjenness Tim Jenness added a comment -

            Apple El Capitan comes with numpy v1.8.0rc1.

            tjenness Tim Jenness added a comment - Apple El Capitan comes with numpy v1.8.0rc1.

            That's a very good question, and possibly related to the discovery that some of my test failure issues come from system python not anaconda.

            >>> import numpy as np
            >>> np.__version__
            '1.8.0rc1'
            

            rhl Robert Lupton added a comment - That's a very good question, and possibly related to the discovery that some of my test failure issues come from system python not anaconda. >>> import numpy as np >>> np.__version__ '1.8.0rc1'
            rowen Russell Owen added a comment -

            rhl reports that casting the image data to double precision (python "float") makes the unit test pass, so this is almost certainly the same issue reported in DM-1984: a known bug in older versions of numpy.

            rowen Russell Owen added a comment - rhl reports that casting the image data to double precision (python "float") makes the unit test pass, so this is almost certainly the same issue reported in DM-1984 : a known bug in older versions of numpy.

            I pushed a trivial fix. Please test it and let me know if it works (as part of reviewing it), since you have an environment that shows the problem.

            rowen Russell Owen added a comment - I pushed a trivial fix. Please test it and let me know if it works (as part of reviewing it), since you have an environment that shows the problem.

            The fix works as advertised. Thank you.

            rhl Robert Lupton added a comment - The fix works as advertised. Thank you.

            People

              rowen Russell Owen
              rhl Robert Lupton
              Robert Lupton
              Robert Lupton, Russell Owen, Tim Jenness
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Jenkins

                  No builds found.