[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Solving A*x=b when A is full rank but numerically rank deficient
|
From: |
withaar |
|
Subject: |
Re: Solving A*x=b when A is full rank but numerically rank deficient |
|
Date: |
Wed, 26 Jun 2013 13:25:25 -0700 (PDT) |
My first effort would be to investigate whether the data can be mapped using
a logarithm. This will make power relations linear and avoid the numeric
instability. Altogether the best if possible.
As a poor second I might use pinv (pseudo-inverse), which produces the
optimal fit in the least-square sense, I believe.
Willem
--
View this message in context:
http://octave.1599824.n4.nabble.com/Solving-A-x-b-when-A-is-full-rank-but-numerically-rank-deficient-tp4653604p4654880.html
Sent from the Octave - General mailing list archive at Nabble.com.