Xref: utzoo comp.lang.c:17869 sci.math.stat:699
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wasatch!cs.utexas.edu!uunet!tektronix!tekcae!kurtk
From: kurtk@tekcae.CAX.TEK.COM (Kurt Krueger)
Newsgroups: comp.lang.c,sci.math.stat
Subject: Re: Need matrix inversion C routine.
Keywords: C matrix invert
Message-ID: <2692@tekcae.CAX.TEK.COM>
Date: 24 Apr 89 15:47:13 GMT
References: <2846@tank.uchicago.edu> <5785@cbnews.ATT.COM> <10087@smoke.BRL.MIL> <12784@pasteur.Berkeley.EDU>
Reply-To: kurtk@tekcae.CAX.TEK.COM (Kurt Krueger)
Followup-To: comp.lang.c
Distribution: usa
Organization: Tektronix, Inc., Beaverton,  OR.
Lines: 19

There are two 'obvious' places to look for code or algorithms:

The Collected Alogrithmns of the ACM  (CACM).  Any technical/college library
	should have this.  The articles usually contain code (but stuff that
	is well known is likely to be in older articles in Fortran or Pascal)
	and references to the algorithmn used.

Any book on numerical methods.

Note:  "C" is not well suited for writing code for dealing with an arbitrary
	2 (or higher) dimension array.  Brush up on your pointers.  You get
	to do the row/column location by hand.  A good source would be old
	Fortran code (Fortran 2 or IBM 1130 Fortran IV) that handles 2-d
	arrays as a strung out 1-d array.

	Another "trick" is to NOT pass a 2-d array, but pass a vector of
	row pointers.  You can then use matrix[row][column] in your function
	as if you have a "real" 2-d array.  But don't expect blinding execution
	speeds.