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

Re: Not a valid limit in NIntegrate ???

• Subject: [mg3144] Re: Not a valid limit in NIntegrate ???
• From: whitic at rpi.edu (whitic)
• Date: 7 Feb 1996 09:38:43 -0600
• Approved: usenet@wri.com
• Distribution: local
• Newsgroups: wri.mathgroup
• Organization: Rensselaer Polytechnic Institute, Troy NY, USA
• Sender: daemon at wri.com

a_kowald at chemie.fu-berlin.de (Axel Kowald) wrote:

>Hi 


>FindRoot[NIntegrate[t,{t,rho,100}]==1,{rho,99}]

>NIntegrate[t,{t,rho,100}] /. rho->99


>When I'm using one of the above constructs either on a Mac or under Unix
>running Mathematica 2.2 I get the following error message:

>NIntegrate::nlim: t=rho is not a valid limit of integration


>Why does it complain ??    Any ideas ??


>Many thanks


>	Axel Kowald

Try using the built in Secant method for the FindRoot routine.
Mathematica uses this method when FindRoot is called with the
following syntax:
	FindRoot[ f[x]==g[x], {x,x1,x2}]

For the problem in question, you can use the command:
	FindRoot[ NIntegrate[t,{t,rho,100}]==1,{rho,99,100}] 
with this, Mathematica does not complain, since it does not have to
evaluate, analytically the Jacobian of this numerical function.

Hope this helps,

Chris Whiting
whitic at rpi.edu