I have to solve Ax=b,conjugate gradient was on my mind,but A is not positive definite,so for a sparse matrix which is the good solution in this case?Gauss Seidel?I do not have to much time, I need to write a program.