MathGroup Archive 2003

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

Search the Archive

Re: NonlinearFit

  • To: mathgroup at smc.vnet.net
  • Subject: [mg42247] Re: NonlinearFit
  • From: lamp02 at hotmail.com (Andrew Kang)
  • Date: Wed, 25 Jun 2003 01:53:42 -0400 (EDT)
  • References: <bd6j0m$ca7$1@smc.vnet.net> <bd8o1p$lo8$1@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

Thanks.

Unfortunately it dose not work.  Could you please give me more advice
on following error message?

Plot::plnr: ft1 is not a machine-size real number at x =
1.25000003125.
Plot::plnr: ft1 is not a machine-size real number at x =
1.2804252436796868.
Plot::plnr: ft1 is not a machine-size real number at x =
1.31360659989453.
General::stop: Further output of Plot :: plnr will be suppressed
during this calculation.

Please e-mail me.

Regards,
Andy

gohtk at rocketmail.com (goh tat kean) wrote in message news:<bd8o1p$lo8$1 at smc.vnet.net>...
> Dear Andy,
> 
> Give a second thought on your problem, I shouldn't multiply your
> function and it's conjugate.
> 
> Just expand the complex function a Exp[I Pi*b/180]/(x - c + I d)^2.5,
> ComplexExpand[a Exp[I Pi*b/180]/(x - c + I d)^2.5, 
>   TargetFunctions -> {Re, Im}]
> 
> You will get something like,
> a Cos[b Pi/180] Cos[2.5 ArcTan[-c + x, d]]/(d^2 + (x - c)^2)^1.25 + (a
> Sin[
>         b Pi/180] Sin[2.5ArcTan[-c + x, d]]/(d^2 + (x - c)^2)^1.25)
> 
> Use this function to do a NonlinearFit on your data.
> 
> Rewrite your function,
> eq[x_, a_, b_, c_, d_] := Re[a Exp[I Pi*b/180]/(x - c + I d)^2.5]
> 
> Generate some dummy data,
> data1 = Table[{x, eq[x, 5, 5, 5, 5]}, {x, 1, 10}]
> 
> Load the NonlinearFit package,
> << Statistics`NonlinearFit`
> 
> Start the fitting,
> ft1 = NonlinearFit[data1, a Cos[b Pi/180] Cos[2.5 ArcTan[-c + x,
> d]]/(d^2 + (x - c)^2)^1.25 + (a Sin[
>         b Pi/180] Sin[2.5ArcTan[-c + x, d]]/(d^2 + (x - c)^2)^1.25),
> x, {a, b, c, d}, ShowProgress -> True]
> 
> 
> 
> 
> 
> Regards,
> tat kean


  • Prev by Date: MathOptimizer Professional software announcement
  • Next by Date: Re: NDSolve documentation
  • Previous by thread: Re: NonlinearFit
  • Next by thread: Wierd ReplaceAll problem