3

Following is the pseudo-code of a simple iterative method of solving $Ax=b$ where $A$ is an $n\times n$ matrix.

for i = 0 .. iteration_count
    for j = 0 .. n
        x = x + A.row(j) * (b[j] - A.row(j) * x) / (A.row(j) * A.row(j)^T)

Note that A.row(j) stands for the j-th row of A.

It seems to be neither Gauss-Seidel nor Jacobi iteration.

Does this method have a name?

badatmath
  • 31
  • (Note: this is a reply to a deleted comment.) x is implied to be set to an arbitrary initial guess x0 at the start. The x's are used as column vectors, so A.row(j) * x represents an inner product or dot product. – badatmath Oct 07 '16 at 04:17

0 Answers0