home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!elroy.jpl.nasa.gov!usc!zaphod.mps.ohio-state.edu!darwin.sura.net!paladin.american.edu!auvm!NIHCU.BITNET!XIC
- Acknowledge-to: XIC@NIHCU.BITNET
- X-Acknowledge-to: XIC@NIHCU.BITNET
- Message-ID: <STAT-L%93010610590200@VM1.MCGILL.CA>
- Newsgroups: bit.listserv.stat-l
- Date: Wed, 6 Jan 1993 10:56:54 EST
- Sender: STATISTICAL CONSULTING <STAT-L@MCGILL1.BITNET>
- From: XIC@NIHCU.BITNET
- Subject: Re: Matrix Problem
- Lines: 19
-
- This is a follow-up to the Matrix Problem of Just Jensen and
- the response by Randy Tobias. If C = Inverse(A), then I = AC.
- Multiplying A by C in block form gives an upper right block
- matrix of A11*C12 + A12*C22 = 0 matrix. The only assumption
- needed is that A is invertible. Pre- and post-multiplying by
- Inverse(A11) and Inverse(C22) respectively (if they exist, and
- they will if A is positive definite) gives Just Jensen's
- formula.
-
- One brief comment on another subject: The Washington
- Statistical Society lectures in the early 1980's on the
- accuracy of statistical packages referred to by Philip
- Gallagher were, I believe, given by Jim Carpenter of the
- Bureau of Labor Statistics and David Morganstein of Westat.
-
- Michael P. Cohen
- National Center for Education Statistics
- Washington, DC 20208-5654
- Bitnet: XIC@NIHCU Internet: XIC@CU.NIH.GOV
-