Hello,
I am becoming acquainted with solving basic linear problems in matlab iteratively. So, I apologize for this very basic question.
I am writing a program to solve Ax=f, for a randomly generated matrix (say, of size 6).
I want to extract to break A into M and N, where A=M-N.
I have written a code for this, and I would like to determine the number of iterations that are required before the solution converges. However, my solution does not converge! It blows up to infinity. I have checked to make sure the matrix is non-singular, and I think I have the iterative scheme correct. I would be very grateful if someone could please check my work and let me know where I might be going wrong. Here is my code:
scale=6;
A = randn(scale,scale);
if det(A)==0
fprintf('Matrix is singular');
else
fprintf('Matrix is non singular');
end
n=size(A,1);
M=tril(triu(A,-1),1);
N=-(M-A);
B=M-N;
f=randn(length(M),1);
x=zeros(length(M),1);
k=1;
r=f-A*x(:,k);
tol=1e-3;
while norm(r)>=tol
x(:,k+1)=M\-N*(x(:,k) + f)
r(:,k)=f-A*x(:,k);
norm(r)
k=k+1;
end
Edit:
Added a condition:
if abs(max(E))>=1
fprintf('Conditions do not hold')
break
end
However, there are times when the random matrix A satisfied this condition, and runs through the while look, and still blows up! Any ideas?
