Exporting to numpy changes the order of dimensions about simpleitk HOT 3 CLOSED

The SimpleITK point is represented as (x,y,z) and simularly the index is (i,j,k). This directly maps to the ITK index and point types. The the sitk Image [] operators can be done img[(i,j,k)] or img[i,j,k]. This provides a consistent interface. Here the x point and i index are the fast axis for the buffer layout.

For numpy's ndarra it is indexed nda[k,j,i] with i being the fast axis. There are ways to manipulate the ndarray image so that the index order is reversed. nda are default indexed by [slice, row, column].

Hope this helps.

from simpleitk.

I got some kind of answer here :
http://insightsoftwareconsortium.github.io/SimpleITK-Notebooks/01_Image_Basics.html

Still not sure why it is that way. Anyway if it's official, I am ok with it. closing

from simpleitk.

Thanks ! good to know

from simpleitk.