MathGroup Archive 2008

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

Search the Archive

Re: Problems with NSolve

  • To: mathgroup at smc.vnet.net
  • Subject: [mg88239] Re: Problems with NSolve
  • From: Gopinath Venkatesan <gopinathv at ou.edu>
  • Date: Tue, 29 Apr 2008 06:49:45 -0400 (EDT)

Kshitij

I had a similar problem sometime back. When I used NSolve[] to solve a set of simultaneous equations (after representing the differentials of the ODE/PDE in terms of function values - just like finite difference scheme), the NSolve[] took sufficiently longer time.

To my surprise, when I used Solve[], I found a big difference in the solution time. Solve[] took me around 30 to 150 seconds (depending on my equations), while for the same set of equations NSolve[] took more than an hour or so to 'running out of memory' error message.

I also thought that setting precision level might help in the set up of NSolve[] (just like you did here), but did not help me.

I also guess (on the similar lines), DSolve[] could solve in less time than NDSolve[] in case of ODE/PDEs.

I wish Mathematica experts give us some good lecture on the use of Solve[] vs NSolve[], DSolve[] and NDSolve[]. Thanks,

Gopinath
University of Oklahoma


  • Prev by Date: Re: Re: Wolfram User Interface Research?
  • Next by Date: Re: Does Mathematica really need more printed,
  • Previous by thread: Re: Re: Re: Problems with NSolve
  • Next by thread: financial chart with volumes