Re: Re: Hold problems with FindRoot

*To*: mathgroup at smc.vnet.net*Subject*: [mg57126] Re: [mg57073] Re: Hold problems with FindRoot*From*: Chris Chiasson <chris.chiasson at gmail.com>*Date*: Mon, 16 May 2005 21:35:24 -0400 (EDT)*References*: <428457A1.3020501@bjornerstedt.org> <200505140858.EAA09467@smc.vnet.net> <acbec1a405051410265918dd00@mail.gmail.com> <4288E5A7.90105@bjornerstedt.org>*Reply-to*: Chris Chiasson <chris.chiasson at gmail.com>*Sender*: owner-wri-mathgroup at wolfram.com

I don't know the answer. I think FindRoot does attempt to symbolically find the derivative of the system with respect to the independant variables. Without knowing exactly what goes on inside of FindRoot, I guess that whatever error it ran into -- when you didn't have the PatternTest -- prevented FindRoot from realizing that it needed to switch to the "secant method" for estimating the derivative. On 5/16/05, Jonas Björnerstedt <lists at bjornerstedt.org> wrote: > Thanks! This solved my problem. Is this an undocumented restriction in > how FindRoot can be used, or where can I find more documentation? > > Jonas > > Chris Chiasson wrote: > > >Try changing the definitions of f and g, via a PatternTest, so that > >they are only defined for numeric arguments. > > > >f[x_?NumericQ,y_?NumericQ]=blahblah1 > >g[x_?NumericQ,y_?NumericQ]=blahblah2 > > > >On 5/14/05, Jonas Björnerstedt <lists at bjornerstedt.org> wrote: > > > > > >>Hello again, > >> > >>There was a small mistake in my previous posting. It should have said: > >> > >>FindRoot[Evaluate[{f[x,y]==0, g[x,y]==0}],{{x,1},{y,1}}] > >> > >>where f and g are functions that work correctly. Trying to solve one > >>equation for two unknowns would cause other problems than I am currently > >>having. > >> > >>Jonas. > >> > >>Jonas Björnerstedt wrote: > >> > >> > >> > >>>Hello all, > >>> > >>>I have problems getting FindRoot to evaluate the equations it > >>>evaluates, despite using Evaluate on the equations. I have a rather > >>>complex function f[x,y] that works correctly with individual values x > >>>and y. Evaluating the following expression does not work however. > >>> > >>>FindRoot[Evaluate[f[x,y]==0],{{x,1},{y,1}}] > >>> > >>>When evaluating, I get similar errors as when I evaluate f without > >>>inserting values for x and y. I can also do a Plot3D on f without > >>>using Evaluate. > >>> > >>>Does anybody have an idea what the problem is? > >>> > >>>Jonas > >>> > >>> > >>> > >> > >> > > > > > > > > > > -- Chris Chiasson http://chrischiasson.com/ 1 (810) 265-3161

**References**:**Re: Hold problems with FindRoot***From:*Jonas Björnerstedt <lists@bjornerstedt.org>