[Insight-developers] Failing itkScalarImageKmeansImageFilter3DTest

[Bill Lorensen]

Luis,

FYI, the kmeans class uses itkWeightedCentroidKdTreeGenerator.

Bill

On Tue, Aug 12, 2008 at 11:51 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
> Hans,
>
> There is a way to set a test-specific timeout in cmake with
> SET_TESTS_PROPERTIES. I just checked in a change.
>
> Bill
>
>
> On Tue, Aug 12, 2008 at 11:37 AM, Hans Johnson <hans-johnson at uiowa.edu> wrote:
>> Bill,
>>
>> This has been assigned a bug http://public.kitware.com/Bug/view.php?id=7481,
>> and Luis has been assigned.  Luis has agreed to look at the bug (see emails
>> at end of this message).
>>
>> Is there a way to set a 60 second timeout for one test only?
>>
>> Hopefully this can be resolved quickly.
>>
>> Hans
>> --
>> Hans J. Johnson, Ph.D.
>> Hans-johnson at uiowa.edu
>>
>> 278 GH
>> The University of Iowa
>> Iowa City, IA 52241
>> (319) 353 8587
>>
>>
>>> From: Bill Lorensen <bill.lorensen at gmail.com>
>>> Date: Tue, 12 Aug 2008 11:29:25 -0400
>>> To: Johnson Hans <hans-johnson at uiowa.edu>
>>> Cc: ITK <insight-developers at itk.org>
>>> Subject: Failing itkScalarImageKmeansImageFilter3DTest
>>>
>>> Hans,
>>>
>>> I see that the test you checked in recently is timing out as you
>>> predicted. What are the plans for fixing this? I understand that this
>>> is a bug, but we should try to fix it ASAP since people get dashboard
>>> e-mails about this test when they check in code. Also, a constantly
>>> failing test may masquerade other failures.
>>>
>>> Bill
>>
>>
>> Hi Hans,
>>
>> Thanks for the heads up.
>>
>> It looks like this is associated to the code modifications
>> related to fixing the KdTree.
>>
>> I'll take a look
>>
>>
>>   Thanks
>>
>>
>>      Luis
>>
>>
>> ----------------------
>> Hans Johnson wrote:
>>> Luis,
>>>
>>> Some code that you changed back in April has surfaced as a problem.  I've
>>> assigned a bug to this issue at
>>> http://public.kitware.com/Bug/view.php?id=7481 and there are now tests that
>>> show how to reproduce the problem.
>>>
>>> It looks like a significant change that was necessary to resolve some other
>>> bug.  I don't want to just roll back the changes without understanding the
>>> problem completely.
>>>
>>> Let us know if there is anything else we can do to help resolve this issue.
>>>
>>> Thanks,
>>> Hans
>>>
>>> ==== pseudo code for identifying the problem. ======
>>>
>>> cd ${INSIGHT_SOURCE}
>>> ## To get the rest of the code in a working form.
>>> cvs update -dAP -D "20080426"
>>>
>>> ## to get the new test added to itk
>>> cd ${INSIGHT_SOURCE}/Testing/Code/Algorithms
>>> cvs update -dAP
>>> ## to get the new test datasets
>>> cd ${INSIGHT_SOURCE}/Examples/Data
>>>
>>> ## First check out version that passes the test
>>> cd ${INSIGHT_SOURCE}
>>> cvs update -r 1.19 Code/Numerics/Statistics/itkStatisticsAlgorithm.txx
>>> cvs update -r 1.9 Code/Numerics/Statistics/itkStatisticsAlgorithm.h
>>>
>>> make -C ${INSIGHT_BUILD_DIR}
>>> pushd ${INSIGHT_BUILD_DIR}
>>> ctest -R itkScalarImageKmeansImageFilter
>>> echo "This test should pass in less than 1 minute"
>>> popd
>>>
>>> ## Now induce the error case that never completes the test
>>> cvs update -r 1.20 Code/Numerics/Statistics/itkStatisticsAlgorithm.txx
>>> cvs update -r 1.10 Code/Numerics/Statistics/itkStatisticsAlgorithm.h
>>>
>>> make -C ${INSIGHT_BUILD_DIR}
>>> pushd ${INSIGHT_BUILD_DIR}
>>> ctest -R itkScalarImageKmeansImageFilter
>>> echo "This test never completes"
>>> popd
>>>
>>
>>
>