Hi,

I had a chance to investigate a bit further and... false alarm. The error was inside our wrapping of Cleaver.

I'm really sorry for the disturbance.

Regards,

Julien.

On Wed, Feb 5, 2014 at 12:18 PM, Jonathan Bronson <bronson@sci.utah.edu> wrote:

Hi Julien

I agree with Josh, we'd need the material volumes you're generating
with that label map in order to diagnose what's going on. Also, as
Josh mentioned I'm on travel, so please excuse any delays in response
this week.

Best,
   Jonathan

On Tue, 4 Feb 2014 17:27:30 -0500
 Joshua Levine <levinej@clemson.edu> wrote:

Hi Julien,

Jonathan is traveling for the moment, so our responses may be a little delayed.

Could you clarify exactly how the NaN's are produced?  Perhaps sending
us the exact data might be required.

I did the following with your code; which did not produce any NaNs:

(1) From the input, I took each material label (0, 17, 143, 209) and
created a binary image
(2) I then ran Cleaver with the 4 input files, this required running
with -p since there were cuts on the boundary.

In this default configuration, everything looks OK.  I did receive a
couple of warnings stating "Failed to Project Triple BACK into 3D.
Using barycenter"

This is with version 1.5.4 of cleaver, freshly downloaded from the
website and built on OSX.

Josh

On Tue, Feb 4, 2014 at 8:34 AM, Julien Finet <julien.finet@kitware.com> wrote:

Hi,

I am using Cleaver 1.5.4 and it works great. Very nice work!
However, with some data, I have NaN coordinates in the generated mesh.
Please see an example volume enclosed. I tried to make it as small as
possible.
There are 4 materials in that volume (0 is air). Let me know if you want the
volume of each label.

Can you please let me know what can be done to fix that ?

Thanks a lot,
Julien.