[Insight-developers] [ITK 0010205]: Recent changes to itkBayesianClassifierImageFilter.txx cause WrapITK failures

Johnson, Hans J hans-johnson at uiowa.edu
Sun Jan 31 12:01:37 EST 2010


Yes.  Will you please roll it back.  I made the cange as a stop gap  
until the  offending bug was fixed.

NOTE:  the class has 4 parameters, but only 3 were wrapped.  The 4 th  
argument always defaulted to double.

Hans

Sent from my iPhone

On Jan 31, 2010, at 8:46 AM, Gaëtan Lehmann  
<gaetan.lehmann at jouy.inra.fr> wrote:

>
> Le 31 janv. 10 à 03:27, Luis Ibanez a écrit :
>
>> Hi Hans,
>>
>> Yes, I would expect that the four template parameters
>> are independent of each other.
>>
>> template <
>>
>>    class TInputVectorImage,
>>    class TLabelsType=unsigned char,
>>    class TPosteriorsPrecisionType=double
>>    class TPriorsPrecisionType=double
>>
>>> BayesianClassifierImageFilter
>>
>>
>> I'm not sure that you actually want to wrap all
>> permutation, but... regardless... the code should
>> compile with different combinations of these
>> template parameters.
>
> Luis, Hans,
>
> Thanks a lot for looking at that problem.
>
> Currently, the following type is wrapped with the default settings:
>
>   itk::BayesianClassifierImageFilter< itk::VectorImage< float,2 >,  
> unsigned short, float >
>
> and the same in 3D.
> Before the recent changes, the following types where wrapped by  
> wrapitk by default
>
>   itk::BayesianClassifierImageFilter< itk::VectorImage< unsigned  
> short,2 >, unsigned short, float >
>   itk::BayesianClassifierImageFilter< itk::VectorImage< float,2 >,  
> unsigned short, float >
>
> and the same in 3D.
>
> I think we should rollback to the previous wrapped types.
> Is it ok for you?
>
> Thanks,
>
> Gaëtan
>
>
>>
>>
>>
>>   Luis
>>
>>
>> --------------------------------------------------------------------
>> On Sat, Jan 30, 2010 at 6:05 PM, Hans Johnson <hans- 
>> johnson at uiowa.edu> wrote:
>>> Luis,
>>>
>>> Are all 4 template parameters completely independent of each  
>>> other?  If so,
>>> then I'll go in and revert the wrapping to do all permutations  
>>> again.
>>>
>>> Hans
>>>
>>>
>>> On 1/30/10 3:30 PM, "Luis Ibanez" <luis.ibanez at kitware.com> wrote:
>>>
>>>> Hi Hans,
>>>
>>> Thanks a lot for pointing this out and for
>>> adding the test case that
>>>> illustrated the problem.
>>>
>>>
>>>> http://public.kitware.com/Bug/view.php?id=10205
>>>
>>> A fix has now been
>>>> committed.
>>>
>>> http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/Algorithms/itkB
>>>> ayesianClassifierImageFilter.txx? 
>>>> root=Insight&r1=1.11&r2=1.12&sortby=date[^]
>>>
>>>>
>>> http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/Algorithms/itkBayesianClas
>>>> sifierImageFilter.h?root=Insight&r1=1.16&r2=1.17&sortby=date
>>> [^]
>>>
>>>
>>> The
>>>> SetPriors() method was calling
>>>
>>>              this->SetInput( 1, priors
>>>> )
>>>
>>> but the SetInput() method is defined in the superclass
>>> (the
>>>> ImageToImageFilter) as a method that takes an
>>> InputImageType * as
>>>> argument.
>>>
>>> That was creating an incorrect entanglement between
>>> the type of the
>>>> Priors and the type of the input image.
>>>
>>>
>>> That call has now been replaced
>>>> with
>>>
>>>  ProcessObject::SetNthInput( 1, priors)
>>>
>>> which expects a DataObject *
>>>> as argument.
>>>
>>>
>>> ---
>>>
>>> BTW: This should also solve one of few
>>> remaining build
>>>> issues in the Dashboard,
>>> that was illustrated in a wrapping build
>>> in
>>>> dash16.
>>>
>>> http://www.cdash.org/CDash/viewBuildError.php?buildid=526705
>>>
>>>
>>>
>>>> Thanks
>>>
>>>
>>>          Luis
>>>
>>>
>>>
>>> ---------- Forwarded message ----------
>>> From:
>>>> Mantis Bug Tracker <mantis at public.kitware.com>
>>> Date: Fri, Jan 29, 2010 at 4:42
>>>> PM
>>> Subject: [ITK 0010205]: Recent changes
>>>> to
>>> itkBayesianClassifierImageFilter.txx cause WrapITK failures
>>> To:
>>>> luis.ibanez at kitware.com
>>>
>>>
>>>
>>> The following issue has been
>>>> ASSIGNED.
>>> ====================================================================
>>>> ==
>>> http://public.kitware.com/Bug/view.php?id=10205
>>> ===========================
>>>> ===========================================
>>> Reported By:
>>>> hans-johnson
>>> Assigned To:
>>>> karthik
>>> === 
>>> ==================================================================
>>>> =
>>> Project:                    ITK
>>> Issue ID:
>>>> 10205
>>> Category:
>>> Reproducibility:            always
>>> Severity:
>>>> major
>>> Priority:                   high
>>> Status:
>>>> assigned
>>> === 
>>> ==================================================================
>>>> =
>>> Date Submitted:             2010-01-29 16:42 EST
>>> Last Modified:
>>>> 2010-01-29 16:42
>>>> EST
>>> === 
>>> ===================================================================
>>> Sum
>>>> mary:                    Recent changes
>>>> to
>>> itkBayesianClassifierImageFilter.txx cause WrapITK
>>>> failures
>>> Description:
>>> Karthik,
>>>
>>> Your recent changes to
>>>> itkBayesianClassifierImageFilter.h have exposed a
>>> template dependancy that was
>>>> causing compile time failures when using
>>> WrapIT
>>> K.  I've posted the compiler
>>>> errors below, and committed code in the
>>> testing directory that can be used to
>>>> force the compile time error.
>>>
>>> It looks like a type mis-match when floating
>>>> point types are used.
>>>
>>> I can make this fail in the
>>>> itkBayesianClassifierImageFilterTest.cxx
>>>
>>> #if 1
>>>   {
>>>   const unsigned int
>>>> TestDimension = 2;
>>>   typedef unsigned char  TestLabelType;
>>>   typedef float
>>>>       TestPosteriorType;
>>>
>>>   //This is a failing test case!!!
>>>   typedef
>>>> float          TestPriorType;·
>>>   typedef itk::VectorImage< double
>>>> ,TestDimension >
>>> TestInitialLabelImageType; //The element type MUST be the
>>>> PriorType
>>>
>>>   typedef itk::BayesianClassifierImageFilter<
>>>
>>>> TestInitialLabelImageType, TestLabelType, TestPosteriorType,
>>> TestPriorType >
>>>>  TestClassifierFilterType;
>>>
>>>> TestClassifierFilterType::Pointer
>>> test=TestClassifierFilterType::New();
>>>
>>>> }
>>> #endif
>>>
>>>
>>> Hans
>>>
>>> ============================================================
>>>> ==========
>>>
>>> Issue History
>>> Date Modified    Username       Field
>>>>
>>>> Change
>>> === 
>>> ===================================================================
>>>>
>>> 2010-01-29 16:42 hans-johnson   New Issue
>>> 2010-01-29 16:42 hans-johnson
>>>> Status                   new => assigned
>>> 2010-01-29 16:42 hans-johnson
>>>> Assigned To               =>
>>>>
>>> karthik
>>> === 
>>> ===================================================================>
>>>
>>>
>>>
>> _______________________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Kitware offers ITK Training Courses, for more information visit:
>> http://kitware.com/products/protraining.html
>>
>> Please keep messages on-topic and check the ITK FAQ at:
>> http://www.itk.org/Wiki/ITK_FAQ
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.itk.org/mailman/listinfo/insight-developers
>
> -- 
> Gaëtan Lehmann
> Biologie du Développement et de la Reproduction
> INRA de Jouy-en-Josas (France)
> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
> http://voxel.jouy.inra.fr  http://www.itk.org
> http://www.mandriva.org  http://www.bepo.fr
>


More information about the Insight-developers mailing list