Services & Resources / Wolfram Forums
-----
 /
MathGroup Archive
2004
*January
*February
*March
*April
*May
*June
*July
*August
*September
*October
*November
*December
*Archive Index
*Ask about this page
*Print this page
*Give us feedback
*Sign up for the Wolfram Insider

MathGroup Archive 2004

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

Search the Archive

Minimization of chi square

  • To: mathgroup at smc.vnet.net
  • Subject: [mg49025] Minimization of chi square
  • From: mfific at ptt.yu (Mario Fific)
  • Date: Tue, 29 Jun 2004 04:49:54 -0400 (EDT)
  • Sender: owner-wri-mathgroup at wolfram.com

Mathematica 5.0

I have one minimization problem: objective function that is minimized
(NMinimize) is chi-square. When try to find minimum first it reports
that there is division by zero, and then returns complex indeterminate
value and stops. I am sure that there is no zero value in denominator
of chi-square, but I guess that the minimization algorithm tries with
some paratmer values that makes the denominator value very close to
zero. My guess is then that NMinimize drop the restriction for one of
my parameters to be integer (I put that constraint) and then that this
make minimization impossible.

Note that when objective function to be minimized is sum of squared
deviation (SSE which is practically the same function as chi-square
but withouth problematic denominator) everything works fine.

I would appreciate any suggestion in order to solve this problem and
use chi-square for minimization.

Mario


  • Prev by Date: Re: one question about draw graph in frame
  • Next by Date: Appropriate codec for playing Mathematica generated .wav files
  • Previous by thread: using updated Intel Math Kernel Library
  • Next by thread: Appropriate codec for playing Mathematica generated .wav files