I thought a householder transformation was only for symmetric matrices.

A househoulder transformation is designed to produce a tridiagonal matrix that has the same eigenvalues as the original matrix. Usually a QR is performed on it afterwards. Anyway, this is a lot of computation and is best done by a computer.

Or did you mean a householder matrix?

]]>A= 4 3 0

4 3 6

2 6 5

This is not covered in my textbook

]]>