Question: Fast calculation of eigenvalues and eigenvectors

I use Maple 10 to calculate eigenvalues and eigenvectors:

K:=Matrix(n,n,datatype=float,storage=sparse):
M:=Matrix(n,n,datatype=float,shape=diagonal):
# code to assign the non-zero elements of K and M
(lambda,v):=Eigenvectors(K,M):
 

For n>1000 the calculation takes several minutes (on 2.4 GHz processor, 2 GB RAM). Maple calculates all eigenvalues, but I only need the ten smallest eigenvalues (and the corresponding eigenvectors). Is there a way to increase the speed of the calculation?

 

 

 

Please Wait...