[Insight-developers] Change in ITK[master]: ENH: Adaptor for vnl Sparse Symmetric Matrix

Tue Jul 9 15:09:19 EDT 2013

Adding Matt McCormic to this e-mail.  and posting on Insight-developers list to recruit more help.

From: Michael Bowers <michaelbowersjhu at gmail.com<mailto:michaelbowersjhu at gmail.com>>
Date: Tuesday, July 9, 2013 2:04 PM
To: Hans Johnson <hans-johnson at uiowa.edu<mailto:hans-johnson at uiowa.edu>>
Subject: Re: Change in ITK[master]: ENH: Adaptor for vnl Sparse Symmetric Matrix

Hans,

I'm really sorry it seems like nothing has been happening on this.  I've been busy and have also run into a number of roadblocks on what I think we both expected would be a pretty simple test to write.

Of course my first instinct was to call the vnl_algo code, then the ITK wrapper code and compare the results.  They differed.  If fact the vnl_sparse_symmetric_eigensystem::CalculateNPairs code returns different results every time I call it for the general case A * x = lambda * B * x.

So I added this routine to vxl.  I naively assumed the way they did the special case A * x = lambda * x would be okay to use, but that works the same way.  They have a static variable that stores the first instantiated variable of this class, and apparently the saved state affects all subsequent calls to CalculateNPairs.

So I'm in a bit of a situation where I don't know what to do.  Would love any suggestions you might have, and of course I'll continue to work the problem.

Michael Bowers
Center for Imaging Science
Johns Hopkins University

On Tue, Apr 23, 2013 at 11:23 AM, Johnson, Hans J <hans-johnson at uiowa.edu<mailto:hans-johnson at uiowa.edu>> wrote:
Thanks!  I'm very excited to have this as part of the ITK toolkit.

Hans

From: Michael Bowers <michaelbowersjhu at gmail.com<mailto:michaelbowersjhu at gmail.com>>
Date: Tuesday, April 23, 2013 10:22 AM
To: Hans Johnson <hans-johnson at uiowa.edu<mailto:hans-johnson at uiowa.edu>>

Subject: Re: Change in ITK[master]: ENH: Adaptor for vnl Sparse Symmetric Matrix

Very good, will do.

Mike B.

On Tue, Apr 23, 2013 at 11:09 AM, Johnson, Hans J <hans-johnson at uiowa.edu<mailto:hans-johnson at uiowa.edu>> wrote:
Michael,

In order for us to add your contribution to ITK, it must be tested.  I added a test that I was 99% sure would fail as a placeholder and reference to what would need to be verified to work on multiple platforms.  I don't know how the test should be written, but I am absolutely confident that a test is needed.

You patch added an entirely new class, but did not provide a test for the class.  This impacts overall code coverage, and more importantly it makes maintenance and stability of ITK difficult.

Please provide some sort of test for your new features so that we can ensure that it works on different compilers/OS/platforms.  At that point we'll be very happy to include your contribution.

Thank you for looking into this,

Hans

From: Michael Bowers <michaelbowersjhu at gmail.com<mailto:michaelbowersjhu at gmail.com>>
Date: Tuesday, April 23, 2013 9:58 AM
To: Hans Johnson <hans-johnson at uiowa.edu<mailto:hans-johnson at uiowa.edu>>
Cc: Arnaud Gelas <arnaudgelas at gmail.com<mailto:arnaudgelas at gmail.com>>
Subject: Re: Change in ITK[master]: ENH: Adaptor for vnl Sparse Symmetric Matrix

Hans,

So I'm trying to figure the status of this stuff.

I added a new class with no unit testing code (VNLSparseSymmetricEigensystemTraits).   Someone (you or Arnaud) added the class to an existing test (itkParameterizationQuadEdgeMeshFilterTest.cxx) and that's now generating compilation errors.  So to go forward I can either fix the current VNL Traits class or write my own test program.  Is that more or less where we are?

The Eigensystem seems pretty different from the Solvers.  Would writing a new test program seem like the wisest course to you at this time?

Thanks for your help,

Mike B.

On Tue, Apr 16, 2013 at 6:47 AM, Hans J. Johnson (Code Review) <review at kitware.com<mailto:review at kitware.com>> wrote:
Hans J. Johnson has posted comments on this change.

Change subject: ENH: Adaptor for vnl Sparse Symmetric Matrix
......................................................................

Patch Set 6: Fails

Updated to match ITK style, and then added a test for the new SSE,  I'm 99% sure that the test is going to fail, but at least there is a placeholder for something to fix.

--
To view, visit http://review.source.kitware.com/10048
To unsubscribe, visit http://review.source.kitware.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I7ef2afb82c500de23085dea2063d4fefb10949b9
Gerrit-PatchSet: 6
Gerrit-Project: ITK
Gerrit-Branch: master
Gerrit-Owner: Michael Bowers <michaelbowersjhu at gmail.com<mailto:michaelbowersjhu at gmail.com>>
Gerrit-Reviewer: Arnaud Gelas <arnaudgelas at gmail.com<mailto:arnaudgelas at gmail.com>>
Gerrit-Reviewer: Hans J. Johnson <hans-johnson at uiowa.edu<mailto:hans-johnson at uiowa.edu>>
Gerrit-Reviewer: Michael Bowers <michaelbowersjhu at gmail.com<mailto:michaelbowersjhu at gmail.com>>

________________________________
Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521, is confidential and may be legally privileged.  If you are not the intended recipient, you are hereby notified that any retention, dissemination, distribution, or copying of this communication is strictly prohibited.  Please reply to the sender that you have received the message in error, then delete it.  Thank you.
________________________________

________________________________
Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521, is confidential and may be legally privileged.  If you are not the intended recipient, you are hereby notified that any retention, dissemination, distribution, or copying of this communication is strictly prohibited.  Please reply to the sender that you have received the message in error, then delete it.  Thank you.
________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-developers/attachments/20130709/b739cb39/attachment.htm>