MD trajectory superposition and transformation

Hi,

I am trying to superpose and transform a set of coordinates(more than 10000, N*3 matrices)to a reference matrix. I can superpose them using superpose function in MD trajectory analysis toolbox but the transformation does not happen. Is Procrustes function in Statistical analysis tool box the right way to do it or what? I appreciate the help.

Answer 1

[Matt J]

If you're comparing 2 sets of Nx3 data sets that differ by an affine transformation then Procrustes might be the thing to use. If you simply have a rototranslation+scaling (i.e., no reflection) there is also this,

http://www.mathworks.com/matlabcentral/fileexchange/26186-absolute-orientation-horns-method

If you have different number of points in each data set and/or it is not a an affine transformation, the situation is more complicated.

[Haleh]

Actually I have more than 10000 sets of data.I was thinking of using Procrustes in a loop. I will try your suggestion and get back to you. Thank you.