MathGroup Archive 1997

[Date Index] [Thread Index] [Author Index]

Search the Archive

Solving linear equations

  • To: mathgroup at smc.vnet.net
  • Subject: [mg9521] Solving linear equations
  • From: Peter <psalzman at landau.ucdavis.edu>
  • Date: Sat, 8 Nov 1997 23:04:47 -0500
  • Sender: owner-wri-mathgroup at wolfram.com

Hi mathgroup

I've got a system of N linear equations (complex coefficients and
variables) and N unknowns.   N is on the order of a few thousand.  I'd
like to know  whether I should use the built in Mathematica function
Solve[] or my own routine.

The catch is--the matrix of equations is *tridiagonal*, and I know of
the Thomas algorithm which efficiently solves this sparse system of
equations.

However, Solve[], being a Mathematica function, is optimized.  Which is
the way to go?  I'm not sure (yet) how close the equations come to
being singular.

Since I'm running Mathematica 3.0.0 (soon to be 3.0.1) on a Pentium II
running at 266MHz, accuracy is more important than speed.

Peter

--
Never criticize a man till you've walked a mile in his shoes.  That way,
you'll be a mile away from him and have his shoes as well.
-=><=-+-+-=><=-+-+-=><=-+-+-=><=-+-+-=><=-+-+-=><=-+-+-=><=-+-+-=><=-+-+-=><=-
Do you hate spam? Join the Coalition Against Unsolicited Commercial
Email (CAUCE) at http://www.cauce.org. Actions speak louder than words.
Join Today!
     I BOYCOTT ANY COMPANY THAT USES MASS ADVERTISING ON THE INTERNET



  • Prev by Date: Re: Write problem with {}
  • Next by Date: Re: cubic splines
  • Previous by thread: RE: variable problems
  • Next by thread: Mathematica and the stock market