home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.theory
- Path: sparky!uunet!mcsun!Germany.EU.net!Urmel.Informatik.RWTH-Aachen.DE!kaa!dak
- From: dak@kaa.informatik.rwth-aachen.de (David Kastrup)
- Subject: Re: Matrix Square Root
- Message-ID: <dak.716054795@kaa>
- Sender: news@Urmel.Informatik.RWTH-Aachen.DE (Newsfiles Owner)
- Nntp-Posting-Host: kaa
- Organization: Rechnerbetrieb Informatik / RWTH Aachen
- References: <Kevin.D.Brunson.1-080992181444@davmac27.oshaughnessy.lab.nd.edu>
- Date: 9 Sep 92 16:06:35 GMT
- Lines: 16
-
- Kevin.D.Brunson.1@nd.edu (Kevin D. Brunson) writes:
-
-
- > This is the first message I have posted to this service so I hope that
- >this is the most appropriate group.
- > The problem I have seems like a simple one: How do I calculate the
- >square root of a symetric matrix? Theoretically, if a matrix A is positive
- >semidefinite then there exists a matrix B that is the square root of A such
- >that A = B*B. I have consulted Lancaster and Tismenetsky's "The Theory of
- >Matrices" but a programmable algorithm is not obvious. The matrix I have
- >generated in monte carlo experiments is 10 x 10 but I could use a smaller
- >one if I could find an operationalization of a method.
- If you do an Eigenvector decomposition A=TDT^{-1}, where D is a diagonal
- matrix containg the eigenvalues of the matrix, and T a matrix containing
- the corresponding eigenvectors in its columns, TD^{0.5}T^{-1} will
- have that property.
-