forwarded message from Brian Gough

forwarded message from Brian Gough

[Brian Gough]

To : Rodney Sparapani <rsparapa at mcw dot edu>
Subject : Re: Missing header in GSL 0.9.3
From : Brian Gough <bjg at network-theory dot co dot uk>
Date : Mon, 8 Oct 2001 09:00:17 +0100 (BST)
References : < 15295.17919.433703.424674@debian ><9B6D336E-BB84-11D5-B173-003065D59DC8@mcw.edu>

Rodney Sparapani writes:
 > I've got this header file in vector:
 > -rw-r--r--  1 rodney  staff   8554 Sep 14 18:16 
 > gsl_vector_complex_long_double.
 > Why it doesn't have a .h extension?  Perhaps, there is a logical limit 
 > to the length of a filename.  I know there is a GNU e-lisp coding 
 > standard that requires files not to exceed 10.3.  Maybe this is why.

Apparently filenames are limited to 31 characters. Not sure what you
can do about that.

forwarded message from Brian Gough

[Brian Gough]

[-- Attachment #0: forwarded message --]
[-- Type: message/rfc822, Size: 1207 bytes --]

From: Brian Gough <bjg@network-theory.co.uk>
To: Andrew Jason Penner <ajpenner@Physics.UManitoba.CA>
Subject: Re: Bug in gsl_linalg_LU_decomp
Date: Sun, 11 Aug 2002 22:53:42 +0100 (BST)
Message-ID: <15702.56550.703139.449129@debian>

Andrew Jason Penner writes:
 > 	Okay, attached is a program that will compute the LU decomposition
 > of a Hermitian matrix
 > 
 > [1 2 3]
 > [2 3 4]
 > [3 4 5]
 > 
 > 	It should fail, however it does produce values in the end. The
 > output is sent to the screen.

The LU decomposition is valid (i.e. if you multiply p'*l*u it gives
the original matrix).

The solver uses the decomposition and gets an answer, it's inaccurate
but there is no division by zero, due to finite precision.

What's needed is an estimate of the error or condition number, which
we currently don't have a function for.  The underlying blas
triangular solvers don't do it.