Failing test on FloodRasterRoadsExperimentalFunction2

[ismailsunni]

Hi @assefay

Would you like to take a look in this failing test:

http://jenkins.inasafe.org/view/QGIS2-InaSAFE-develop/job/safe-package/lastCompletedBuild/testReport/safe.impact_functions.inundation.test_qgis_native_impact_functions/Test_gis_native_impact_functions/test_raster_roads_impact_optimized/

Error Message

21 != 25
-------------------- >> begin captured logging << --------------------
InaSAFE: DEBUG: calculate_impact called with:
Layers: [<safe.impact_functions.inundation.test_qgis_native_impact_functions.Wrapper instance at 0x70006c8>, <safe.impact_functions.inundation.test_qgis_native_impact_functions.Wrapper instance at 0x7000998>]
Function:<class 'safe.impact_functions.inundation.flood_raster_roads_experimental-optimized.FloodRasterRoadsExperimentalFunction2'>
--------------------- >> end captured logging << ---------------------
Stacktrace

  File "/usr/lib/python2.7/unittest/case.py", line 327, in run
    testMethod()
  File "/var/lib/jenkins/inasafe-dev/safe/impact_functions/inundation/test_qgis_native_impact_functions.py", line 357, in test_raster_roads_impact_optimized
    self.assertEquals(count, 25)
  File "/usr/lib/python2.7/unittest/case.py", line 511, in assertEqual
    assertion_func(first, second, msg=msg)
  File "/usr/lib/python2.7/unittest/case.py", line 504, in _baseAssertEqual
    raise self.failureException(msg)
21 != 25
-------------------- >> begin captured logging << --------------------
InaSAFE: DEBUG: calculate_impact called with:
Layers: [<safe.impact_functions.inundation.test_qgis_native_impact_functions.Wrapper instance at 0x70006c8>, <safe.impact_functions.inundation.test_qgis_native_impact_functions.Wrapper instance at 0x7000998>]
Function:<class 'safe.impact_functions.inundation.flood_raster_roads_experimental-optimized.FloodRasterRoadsExperimentalFunction2'>

[ismailsunni]

I only updated the unit test, so this test pass based on what I saw in the QGIS.

[assefay]

@ismailsunni sorry for the delay. looking at it now ...

[ismailsunni]

@assefay I made an update to the unit test (the expected value) based on what I got from QGIS. I hope it's right.

[assefay]

@ismailsunni I think there seem to be a problem with the function it's self. The results should not have changed. I am looking into the steps to see what has caused the changes in the result.

[ismailsunni]

Hmm, ok.

So, my fix is not right. We can revert my fix.

[assefay]

@ismailsunni , @timlinux

The function was "broken" at one point (a shift of about 1/2 pixel in the corners). This was creating the failure in the test. I tried to figure this out yesterday but missed the deadline. I just fixed it now and will be committing. all test case are now ok. I reverted back the value in the test case to what it was originally. Thanks @ismailsunni. And sorry for the delay.

[timlinux]

Hi @assefa

On Wed, Jun 11, 2014 at 10:31 PM, Yewondwossen Assefa < notifications@github.com> wrote:

@ismailsunni https://github.com/ismailsunni , @timlinux https://github.com/timlinux

The function was "broken" at one point (a shift of about 1/2 pixel in the corners). This was creating the failure in the test. I tried to figure this out yesterday but missed the deadline. I just fixed it now and will be committing. all test case are now ok. I reverted back the value in the test case to what it was originally. Thanks @ismailsunni https://github.com/ismailsunni. And sorry for the delay.

Thanks for this. We found a few tests failing under windows so there is still time for some last minute tweaks today if needed.

Regards

Tim

— Reply to this email directly or view it on GitHub https://github.com/AIFDR/inasafe/issues/1030#issuecomment-45757365.

Tim Sutton

Visit http://linfiniti.com to find out about:

• QGIS programming services
• GeoDjango web development
• QGIS Training

• FOSS Consulting Services Skype: timlinux Irc: timlinux on #qgis at freenode.net Tim is a member of the QGIS Project Steering Committee