<html><head><meta http-equiv="Content-Type" content="text/html charset=utf-8"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class="">Just to follow up my git bisect narrowed the change down to this:<div class=""><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; color: rgb(175, 173, 36);" class="">commit 0fd462a3226e84bdce77b68830903b3888dab244</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">Author: Hyun Jae Kang <<a href="mailto:hyunjae.kang@kitware.com" class="">hyunjae.kang@kitware.com</a>></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">Date:   Thu Dec 31 10:56:25 2015 -0500</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; min-height: 13px;" class=""><br class=""></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">    BUG: Fixed the runtime crash of ITKStatisticsTestDriver tests on OSX 10.6</div><p style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; min-height: 13px;" class="">    <br class="webkit-block-placeholder"></p><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">    - Fixed the runtime crash of the following ITKStatisticsTestDriver's tests on</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">      OSX 10.6 ( clang 3.0 )  by modifying the itkHistogram's Initialize function</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">      to handle a data value out of range of a specific data type with</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">      type-casting operator.</div><p style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; min-height: 13px;" class="">    <br class="webkit-block-placeholder"></p><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">      - itkSampleToHistogramFilterTest3 (ILLEGAL)</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">      - itkSampleToHistogramFilterTest7 (ILLEGAL)</div><p style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; min-height: 13px;" class="">    <br class="webkit-block-placeholder"></p><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">    - These crash were reported at the following page:</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">      <a href="https://open.cdash.org/viewTest.php?onlyfailed&buildid=4170483" class="">https://open.cdash.org/viewTest.php?onlyfailed&buildid=4170483</a></div><p style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; min-height: 13px;" class="">    <br class="webkit-block-placeholder"></p><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">    Change-Id: I2063912edee79422d88125ffd5f7513c31a9e98c</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; min-height: 13px;" class=""><br class=""></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class=""><b class="">diff --git a/Modules/Numerics/Statistics/include/itkHistogram.hxx b/Modules/Numerics/Statistics/include/itkHistogram.hxx</b></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class=""><b class="">index 380b8d0..d11d788 100644</b></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class=""><b class="">--- a/Modules/Numerics/Statistics/include/itkHistogram.hxx</b></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class=""><b class="">+++ b/Modules/Numerics/Statistics/include/itkHistogram.hxx</b></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class=""><span style="font-variant-ligatures: no-common-ligatures; color: #34bbc7" class="">@@ -248,8 +248,8 @@</span> Histogram< TMeasurement, TFrequencyContainer ></div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">     {</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">     if ( size[i] > 0 )</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo;" class="">       {</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; color: rgb(195, 55, 32);" class="">-      interval = static_cast<float>( upperBound[i] - lowerBound[i] )</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; color: rgb(195, 55, 32);" class="">-        / static_cast< MeasurementType >( size[i] );</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; color: rgb(52, 189, 38);" class="">+      interval = (static_cast<float>( upperBound[i] ) - static_cast<float>(lowerBound[i] ))</div><div style="margin: 0px; font-size: 11px; line-height: normal; font-family: Menlo; color: rgb(52, 189, 38);" class="">+        / static_cast< float >( size[i] );</div><div class=""><br class=""></div><div class=""><br class=""></div><div class="">Here is the failing example [1], which uses the HistogramMatching filter.</div><div class=""><br class=""></div><div class="">I am not sure why float was choose at the type here. Are results being truncated now? When there was temporary double computation going on?</div><div class=""><br class=""></div><div class=""> It needs more investigation, as it’s changing the results of some filters… not sure how significant yet.</div><div class=""><br class=""></div><div class="">Brad</div><div class=""><br class=""></div><div class=""><br class=""></div><div class=""><br class=""></div><div class=""><br class=""></div><div class="">[1] <a href="https://github.com/SimpleITK/SimpleITK/blob/master/Examples/DemonsRegistration1.py#L41" class="">https://github.com/SimpleITK/SimpleITK/blob/master/Examples/DemonsRegistration1.py#L41</a></div><div><blockquote type="cite" class=""><div class="">On Jan 16, 2016, at 2:58 PM, Bradley Lowekamp <<a href="mailto:brad@lowekamp.net" class="">brad@lowekamp.net</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><div class="">Thanks. I'll run a git bisect tonight to try to figure it out. <br class=""><br class=""><blockquote type="cite" class="">On Jan 16, 2016, at 12:57 PM, Matt McCormick <<a href="mailto:matt.mccormick@kitware.com" class="">matt.mccormick@kitware.com</a>> wrote:<br class=""><br class="">Hi,<br class=""><br class="">Nothing comes to mind.<br class=""><br class="">Here is the log for v4.9rc03..release:<br class=""><br class="">Bill Hoffman (3):<br class="">     COMP: fix 64 bit build warnings with windows auto-export on.<br class="">     COMP: work around for VS 2015 optimizer bug causing test failures.<br class="">     COMP: work around for VS 2015 optimizer bug causing test failures.<br class=""><br class="">Bradley Lowekamp (1):<br class="">     BUG: Fix precision with accumulation and scaling in AdaptiveHistogram<br class=""><br class="">Davis Vigneault (1):<br class="">     COMP: Include itkMacro.h in itkTestingMacros.h<br class=""><br class="">Hans Johnson (2):<br class="">     COMP: BSD command lines do not have --version<br class="">     COMP: Respect CMAKE_CXX_STANDARD during config<br class=""><br class="">Hyun Jae Kang (12):<br class="">     COMP: Fixed the compiler error of ITKCommon2TestDriver on OSX 10.6<br class="">     BUG: Fixed the runtime crash of vnl_test_complex on OSX 10.6<br class="">     BUG: Fixed the runtime crash of itkTimeProbeTest2<br class="">     BUG: Fixed the runtime crash of VideoSourceTest on OSX 10.6<br class="">     BUG: Fixed the runtime crash of ITKReviewTestDriver on OSX 10.6<br class="">     BUG: Fixed the runtime crash of ITKFastMarchingTestDriver's<br class="">tests on OSX 10.6<br class="">     COMP: Fixed the data conversion warning messages of itkResourceProbe<br class="">     BUG: Fixed the runtime crash of ITKStatisticsTestDriver tests on OSX 10.6<br class="">     BUG: Fixed the runtime crash of itkBinaryShapeOpeningImageFilterTest1<br class="">     BUG: Fixed the runtime crash of test_pow_log on OSX 10.6<br class="">     BUG: Fixed the runtime crash of vnl_test_numeric_traits<br class="">     BUG: Exclude a test code of ITKLabelMapTestDriver on OSX 10.6<br class=""><br class="">Isaiah Norton (1):<br class="">     COMP: fix build with -std=c++11 when ITK/VXL initially configured without<br class=""><br class="">Matthew McCormick (6):<br class="">     BUG: Do not perform dynamic_cast in CompositeIOTransformIOHelper.<br class="">     COMP: Do not set property on itkhdf5 with ITK_USE_SYSTEM_HDF5.<br class="">     DOC: Update the location of the GNUPlot Software Guide scripts.<br class="">     DOC: Update the Software Guide repo location in Examples/README.txt.<br class="">     COMP: Add export specification for itk::ResourceProbe.<br class="">     COMP: Require NO_MODULE with DCMTK find_package.<br class=""><br class="">Michka Popoff (2):<br class="">     ENH: Allow Python 3 wrapping for Glue and Review Modules with VTK 7<br class="">     COMP: Re-enable review module for python 3 and older VTK's<br class=""><br class="">Sean McBride (1):<br class="">     DOC: Update some woefully out-of-date GDCM comments, and typos<br class=""><br class="">Seun Odutola (1):<br class="">     BUG: Improved itkAnalyzeImageIO to handle case insensitive extensions<br class=""><br class="">HTH,<br class="">Matt<br class=""><br class=""><blockquote type="cite" class="">On Sat, Jan 16, 2016 at 12:18 PM, Bradley Lowekamp <<a href="mailto:brad@lowekamp.net" class="">brad@lowekamp.net</a>> wrote:<br class="">I just updated SimpleITK to the release branch yesterday and there are some changes in the result of the demons registration. I'm not going to get a chance to look at it until Monday. The only change potentially related I know of is the rounding related numeric traits, but that does not seem likely to me.<br class=""><br class="">Any ideas why the Demons Registration would have changed fro rc3?<br class=""><br class=""><blockquote type="cite" class="">On Jan 16, 2016, at 12:13 PM, Bradley Lowekamp <<a href="mailto:brad@lowekamp.net" class="">brad@lowekamp.net</a>> wrote:<br class=""><br class="">Hello,<br class=""><br class="">I updated SimpleITK<br class="">_______________________________________________<br class="">Powered by <a href="http://www.kitware.com" class="">www.kitware.com</a><br class=""><br class="">Visit other Kitware open-source projects at<br class=""><a href="http://www.kitware.com/opensource/opensource.html" class="">http://www.kitware.com/opensource/opensource.html</a><br class=""><br class="">Kitware offers ITK Training Courses, for more information visit:<br class="">http://kitware.com/products/protraining.php<br class=""><br class="">Please keep messages on-topic and check the ITK FAQ at:<br class="">http://www.itk.org/Wiki/ITK_FAQ<br class=""><br class="">Follow this link to subscribe/unsubscribe:<br class="">http://public.kitware.com/mailman/listinfo/insight-developers<br class=""></blockquote>_______________________________________________<br class="">Powered by <a href="http://www.kitware.com" class="">www.kitware.com</a><br class=""><br class="">Visit other Kitware open-source projects at<br class=""><a href="http://www.kitware.com/opensource/opensource.html" class="">http://www.kitware.com/opensource/opensource.html</a><br class=""><br class="">Kitware offers ITK Training Courses, for more information visit:<br class="">http://kitware.com/products/protraining.php<br class=""><br class="">Please keep messages on-topic and check the ITK FAQ at:<br class="">http://www.itk.org/Wiki/ITK_FAQ<br class=""><br class="">Follow this link to subscribe/unsubscribe:<br class="">http://public.kitware.com/mailman/listinfo/insight-developers<br class=""></blockquote></blockquote>_______________________________________________<br class="">Powered by <a href="http://www.kitware.com" class="">www.kitware.com</a><br class=""><br class="">Visit other Kitware open-source projects at<br class=""><a href="http://www.kitware.com/opensource/opensource.html" class="">http://www.kitware.com/opensource/opensource.html</a><br class=""><br class="">Kitware offers ITK Training Courses, for more information visit:<br class="">http://kitware.com/products/protraining.php<br class=""><br class="">Please keep messages on-topic and check the ITK FAQ at:<br class="">http://www.itk.org/Wiki/ITK_FAQ<br class=""><br class="">Follow this link to subscribe/unsubscribe:<br class="">http://public.kitware.com/mailman/listinfo/insight-developers<br class="">_______________________________________________<br class="">Community mailing list<br class="">Community@itk.org<br class="">http://public.kitware.com/mailman/listinfo/community<br class=""></div></div></blockquote></div><br class=""></div></body></html>