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

Re: Re: Problem with NMaximize

• To: mathgroup at smc.vnet.net
• Subject: [mg90103] Re: [mg90094] Re: Problem with NMaximize
• From: Andrzej Kozlowski <akoz at mimuw.edu.pl>
• Date: Sun, 29 Jun 2008 05:34:33 -0400 (EDT)
• References: <200806280954.FAA26321@smc.vnet.net>

NSolve most certainly does not use a  secant search or other iterative  
procedure (FindRoot does). For a single polynomial equation it uses  
the eigenvalues method, for systems of polynomial equations numerical  
Groebner basis.

Andrzej Kozlowski


On 28 Jun 2008, at 18:54, Bill Rowe wrote:

> On 6/27/08 at 6:16 AM, drmajorbob at att.net (DrMajorBob) wrote:
>
>> NSolve and NMaximize generally use a kind of secant search, using
>> previous attempts to extrapolate to a new and (hopefully) better
>> one.
>
> I do not know what algorithms NSolve uses. But the documentation
> for NMinimize and NMaximize indicate the algorithms used are not
> well described as a "kind of secant search, using previous
> attempts to extrapolate"
>
> See:
>
> tutorial/ConstrainedOptimizationOverview
>

• References:
  • Re: Problem with NMaximize
    • From: Bill Rowe <readnews@sbcglobal.net>