Re: eigenvalue/eigenvector question again

From: Dave (dave_and_darla_at_Juno.com)
Date: 02/26/05 

Date: 26 Feb 2005 11:54:29 -0800

D is the diagonal matrix of eigenvalues of A, so in the absence of
rounding errors in computing A and then computing the eigenvalues of A,
you could recover A.

However, there is no way to recover U exactly, though, because if A =
UDU' and E is any diagonal matrix with diagonal entries = +1 or -1,
with an even number of -1 entries, the A = (UE)D(EU') and det(EU) = 1,
so UE also is a solution.

Dave

Relevant Pages

  • Re: generating random matrix with negative eigenvalues
    ... With D a diagonal matrix and V a real unitary matrix (as generated by ... and its eigenvalues will be the same as those of D, ... the algorithm proceeds. ... If you begin with a random ordering of the ...
    (comp.soft-sys.matlab)
  • Re: The cosine of a matrix
    ... Lancaster. ... I have a related question, though: is it true that for a matrix A we ... it sums the (diagonal matrix of) eigenvalues to 1 and leads ...
    (sci.math)
  • Re: The cosine of a matrix
    ... I have a related question, though: is it true that for a matrix A we ... it sums the (diagonal matrix of) eigenvalues to 1 and leads ... do not know whether it is also valid for defective matrices (I ... The original poster ignores the possibility where the two eigenvalues are ...
    (sci.math)
  • Re: symmetric SVD A=QSQ^T with Q orthonormal?
    ... SVD). ... entries of S are eigenvalues of A; the columns of Q are eigenvectors. ... then A = Q D Q^T where D is a diagonal matrix whose diagonal entries ...
    (sci.math)