MathGroup Archive 2002

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

Search the Archive

Re: Plot[x Sin[x],{-100,100}] Bad {-100,99} Good?

  • To: mathgroup at smc.vnet.net
  • Subject: [mg38313] Re: Plot[x Sin[x],{-100,100}] Bad {-100,99} Good?
  • From: Tom Burton <tburton at brahea.com>
  • Date: Thu, 12 Dec 2002 01:32:41 -0500 (EST)
  • Sender: owner-wri-mathgroup at wolfram.com

On 12/10/02 1:46 AM, in article at4d54$etu$1 at smc.vnet.net, "Steven T.
Hatton" <hattons at globalsymmetry.com> wrote:

> This is Mathematica 4.2 on Linux.  I evaluated Plot[x Sin[x], {-100,100}] and
> disvovered that the portion of the plot between,9.5Pi and 11Pi is displayed
> as a line segment joining the points {9.5 , 9.5 Sin[9.5]} and {11 , 11
> Sin[11]}.  I've also found that the plot from x=-100 to x=99.99 is incorrect.
> The plot over the range [-100,99] is correct.  What is happening here?  Does
> this happen on other systems?
> -- 

Although Mathematica attempts to refine the plot recursively as needed to
produce an accurate plot, you can fool the algorithm. One way is to provide
a function with more oscillations than the initial number of PlotPoints. On
my system (4.2 on Mac), the default initial number of PlotPoints is 25. Your
function oscillates about 32 times over the domain. Increase the number of
PlotPoints to 50. The problem goes away.

As you have seen, fewer PlotPoints may work, depending upon the initial
distribution of points. Plot your examples with PlotDivision->1 to see that
small adjustments in the domain change the number of oscillations captured
by the initial set of points.

Tom Burton



  • Prev by Date: Re: The number of solutions to n_1 + n_2 + n_3 + ... + n_k = m
  • Next by Date: Re: executing 'ProductLog' from an external program (C code or shell)
  • Previous by thread: RE: Plot[x Sin[x],{-100,100}] Bad {-100,99} Good?
  • Next by thread: Re: Plot[x Sin[x],{-100,100}] Bad {-100,99} Good?