Processing math: 100%

 

 

 

Eigenvalues with the QR algorithm and Lanczos' method

We have now found ˆQ and ˆU and this allows us to find the matrix ˆB which is, due to Schur's theorem, unitarily similar to a triangular matrix (upper in our case) since we have that ˆQ1ˆAˆQ=ˆB, from Schur's theorem the matrix ˆB is triangular and the eigenvalues the same as those of ˆA and are given by the diagonal matrix elements of ˆB. Why?

Our matrix ˆB=ˆUˆQ.