Re: orthogonal polynomials relative to measures with mass point
- From: carlos@xxxxxxxxxxxx
- Date: 19 Aug 2005 09:49:50 -0700
Then quad precision (if you are working on a 64-bit processor) may
help.
Failing that, there are improved orthogonalization techniques: Modified
Gram-Schmidt (MGS) instead of GS, and Kahan's theorem for MGS.
My source there is Parlett's The Symmetric Eigenvalue Problem, sec 6.9:
"orthogonalization in the face of roundoff" This classic has been
recently
reprinted by SIAM.
.
- References:
- orthogonal polynomials relative to measures with mass point
- From: Gert Van den Eynde
- Re: orthogonal polynomials relative to measures with mass point
- From: carlos
- Re: orthogonal polynomials relative to measures with mass point
- From: Gert Van den Eynde
- orthogonal polynomials relative to measures with mass point
- Prev by Date: Re: orthogonal polynomials relative to measures with mass point
- Next by Date: Re: Eigenvalues/eigenvectors for a matrix of the form (A^T)A
- Previous by thread: Re: orthogonal polynomials relative to measures with mass point
- Next by thread: Inverse Neumann boundary condition
- Index(es):