Hi everyone and a happy new year,

I am trying to analyse the shape correlation of a structure with age. So far, I've run the SPHARM-PDM / MANCOVA pipeline successfully.

I would like to see what I get without the SPHARM-PDM parameterization.
For this purpose I aligned all of the surfaces with Procrustes alignment (rigid + affine) and projected the differences to a template using
      MeshMath template surface_n.txt -subtract surface_n
as suggested in a previous thread.
Then I use the surface_n.txt files as input for the shape mancova using the following command:
      shapeAnalysisMANCOVA data.csv  --numGroupTypes 1 --numIndependent 1 --interactionTest --simpleCorrs --debug --KWMinput

The command works up to the point that is about to run the simple correlation testing.
Then I get a segmentation fault at line 28 of nonMancova.cxx
       tempNorm=ArrayNormal->GetTuple(feat);

which suggests a problem with the meanSurfaceNormals.
I have downloaded/compiled the version of mancova in http://www.insight-journal.org/browse/pu...

Could you help me with that?