Prev: deleting rows from matrices
Next: Cell array Conversion
From: Dushyant Kumar on 13 Aug 2010 15:42
I am solving an ill-defined linear equation

A_Spatial_Comp * x2_1D_SimultReg = y_exp_1D_Spatial
Under constraint:
Where x2_1D_SimultReg > 0

“A_Spatial_Comp” is a sparse matrix where only the diagonal term is nonzero with diagonal block as 50 x 80.

I am using
[x2_1D_SimultReg,resnorm,residual,exitflag] = lsqlin(A_Spatial_Comp, y_exp_1D_Spatial,A1,b,[ ],[ ],lb,ub,x2_1D_NoRegul,options);

If I try to solve the equation for entire voxels 256 x 128, then the elements of A_Spatial_Comp becomes huge ( 4259840, 2654208). I can divide this problem in smaller chunks so that time in solving these is minimized. I do need suggestion regarding how to split this problem optimally.

Regards,
DK
 | 
Pages: 1
Prev: deleting rows from matrices
Next: Cell array Conversion