Matlab solution to Ax=b fails

I have encountered the following problem: I need to solve a large linear system Ax=b, with complex, square A. I have imported A and b into Matlab from a file, and they are in Matlab's sparse matrix format. I then write x = A/b; and Matlab produces the solution for x. Now I check that the answer x satisfies the original equation: norm(A*x-transpose(b)); The transpose() has been applied to b because as far as I can tell Matlab returns x opposite to the mathematical convention (i.e. a row vector rather than a column vector).

When I run this check, the residual (norm of the difference) is on the order of 10^3, indicating that the solution is completely incorrect. Why is Matlab giving the wrong answer?

Any help would be appreciated, thank you.

Answer 1

[Steven Lord]

A/b ( MRDIVIDE ) and A\b ( MLDIVIDE ) are two different operations.

[Daniel Walsh]

Yes, this was exactly the problem. Thank you for pointing that out.