On Oct 3, 11:08 pm, "Anusha " <anu...@cs.usm.my> wrote: > TideMan <mul...@gmail.com> wrote in message <firstname.lastname@example.org>... > > On Thursday, October 4, 2012 3:45:12 PM UTC+13, Anusha wrote: > > > Hi, > > > > I have a mean point set, represented by 3D coordinates(x,y,z) in 3D matrix. There a another point set in 3D coordinates. How can I compute the standard deviation between the mean and another point set? I want to know how much the dispersion of the given point set from the mean point set. > > > > Thanks. > > > Not sure I understand exactly what you're getting at, but how about: > > std(S1(:)-S2(:)); > > where S1 and S2 are the 3D matrices. > > But S1 and S2 have different number of size. Its like S1 is Mx3 and S2 is Nx3.
This is just nearest mean clustering.
S1 are M cluster means S2 are N points to be assigned to clusters
So all you have to do is assign each member of S2 to the closest member of S1.