I have a conceptual question. The orthogonal shape modes in shape space, and I understand that this space is dimension N (number of correspondences). How are the PCA vectors for each point related to these orthogonal vectors. Is there a way to reconstruct the orthogonal vectors from the 3D vectors, or to get them out of ShapeWorks?
By “orthogonal shape modes”, do you mean the PCA modes of variation?
You can export the eigenvectors in Studio. These are the vectors that the modes of variation show in Studio. A given shape (set of correspondences) can be reconstructed from the mean shape and a single value for each eigenvector. A row of 0’s would be the mean shape. Setting the first value to SD+2 (of the population values for) and the rest to 0 will give a correspondence for mode 0, SD+2.
The raw component scores for each population sample can be exported with “Export Raw Component Scores”. This will be one number for each eigenvector and combined with the mean shape can reconstruct each original population sample.
Does that answer your question?