[Insight-developers] Re: Tests failing after checkins

[Luis Ibanez]

This is a multi-part message in MIME format.
--------------070508010508000100070408
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit


Some more in itkThresholdSegmentationLevelSetImageFilterTest...

The problems seems to be related the convergence of the method.
Attached are the output of the RMS error when the test was passing
in ct02.crd, and the output when the test is failing.

The tests has a maximum number of 1000 iterations. It seems that
depending on the load in the machines the time is sometimes
enough to complete the test before being killed.

The stopping criterion is to have an RMS < 0.02 (blue line on the
attached graphic).  Before friday changes, the test was reaching this 
threshold in 129 iterations.

Now, the RMS errors keeps oscilating around 0.1 until the 1000
iterations have passed.

I'm still tracking what in the changes could have introduced this
error...


Luis


===========================================

Luis Ibanez wrote:

> Hi Bill,
> 
> I'll look at the failing tests:
> 
> - itkCannyEdgeDetectionImageFilterTest
> - itkThresholdSegmentationLevelSetImageFilterTest
> 
> 
> They passed on my Linux box though...
> 
> 
> Luis
> 
> 
> 
> ==================
> Lorensen wrote:
>  >
> 
>> Luis,
>> I noticed that two tests started failing after the checkins yesterday. 
>> One is crashing and the other times out.
>>
>> Bill
>>
>>
>>
> 
> 
> _______________________________________________
> Insight-developers mailing list
> Insight-developers@public.kitware.com
> http://public.kitware.com/mailman/listinfo/insight-developers
> 



--------------070508010508000100070408
Content-Type: image/png;
 name="ThresholdOutput.png"
Content-Transfer-Encoding: base64
Content-Disposition: inline;
 filename="ThresholdOutput.png"
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--------------070508010508000100070408--