Re: Preconditioning and weighted least squares

From: Ray Koopman (koopman_at_sfu.ca)
Date: 03/17/05

  • Next message: Tom: "Re: Confused by FFTW output" 
    Date: 16 Mar 2005 16:53:51 -0800

    spasmous wrote:
    >
    > I've seen another popular type of preconditioner is to transform
    >
    > A.x = b
    >
    > into
    >
    > A.P.x = b
    >
    > This one looks statistically OK to me (ie. we are not weighting the
    > measurements in an arbitrary way).

    Yes. If you solve A.x = b for x, and A.P.y = b for y,
    then you should have P.y = x, to within roundoff.

  • Next message: Tom: "Re: Confused by FFTW output"