Re: Solving functional nonlinear equation

[Jason Martin]

Chang,

    You might want to try the command backend("fltk") .  I noticed a 
long delay while using GNUPLOT with some of my graphing as well, and 
this was one of the solutions that worked for me.  I am putting the 
email at the end of this email in case you missed it previously.  I 
tried it with the sample code below, cleaned up cause I was paranoid 
about the denominator subtractions:

x=linspace(.1,1,20);y=linspace(.1,1,20);[xx,yy]=meshgrid(x,y);
for a=[1.1,1.2,1.5,2]
    Hfxy=@(x,y)( 
(-1/(a^2))*(x.*y).^(1/(a-1))+(1/a)*x.^(1/(a-1))+(1/a)*y.^(1/(a-1)));
    Dfxy=Hfxy(xx,yy);
    figure;surf(xx,yy,Dfxy);shading interp;fflush(1);
end


and it took almost no time for the plots to be created by fltk.

Jason

On 12/28/2009 9:55 AM, Chengqi Chang wrote:
> Please excuse me for any confusing statement in my previous problem.
> My problem is: if data of points is huge, say 100 grid points along both x
> and y axes, my gnuplot renders the figure slower than I can tolerate. I
> think it is the problem with gnuplot while not with octave.
> Thank everyone all the same.
>
>    
> > BTW: when I firstly divided both x,y into 100 grid points, GNU plot
> > functions very very slow(Octave 3.2.0, GNU plot 4.3.0). I much appreciate
> > you help if someone can tell me how to improve its performance.
> >
> >      
>
>    

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

On 12/21/2009 8:24 PM, martin_helm wrote:

> >
> >
> >  Jason Martin-13 wrote:
> >  
>    
> > >>  Good afternoon list,
> > >>
> > >>  Please forgive me this one time error should this have been asked
> > >>  before, or this is in the wrong list.  If it is both, then I guess I am
> > >>  screwed. o.0
> > >>
> > >>  I am trying to figure out how to get clearer plots out of Octave.  I am
> > >>  using this m-file{http://www.thumbprintpro.com/math/fib.m}, and it is
> > >>  giving me the following output:
> > >>  http://www.thumbprintpro.com/math/OctaveOutPut.png  .  While this is
> > >>  adequate, it is difficult to see that this is a 3d image, and compared
> > >>  to this output from Matlab,
> > >>  http://www.thumbprintpro.com/math/MatlabOutPut.png, it is easy to see my
> > >>  frustration.  I am trying to relieve my stress of having to log into my
> > >>  schools slow remote desktops and do all of my math computations through
> > >>  Octave, but the output is of such quality that I cannot insert that into
> > >>  a report.  Also, it took 15 minutes to generate that image where Matlab
> > >>  took 15 seconds.  I have been looking online for a replacement plotter
> > >>  for use with Octave, but nothing I have found yet is working with
> > >>  Windows.  It could just be that I lack the necessary skills to compile
> > >>  some of these packages, or the knowledge of how to insert them into
> > >>  Octave, I am unsure.  Can someone point me in the right direction to
> > >>  getting opengl plots for Octave under Windows?
> > >>
> > >>  Jason
> > >>  _______________________________________________
> > >>  Help-octave mailing list
> > >>  address@hidden
> > >>  https://www-old.cae.wisc.edu/mailman/listinfo/help-octave
> > >>
> > >>
> > >>  
> >      
> >  To get a somewhat improved 3d feeling without the need to install an
> >  additional package
> >  you can use
> >  backend("fltk")
> >  before you call your script and then substitute your simple isosurface call
> >  to
> >
> >  [F, v, c] = isosurface (x, y, z, f, 1.1, x);
> >  p = patch ("Faces", F, "Vertices", v, "FaceVertexCData", c, \
> >      "FaceColor", "interp", "EdgeColor", "none");
> >  colormap(copper(256));
> >  shading interp
> >  view(30,30)
> >
> >  this adds some shading in x direction.
> >
> >  Btw, jhandles gives at the moment the best results, because you can set
> >  lights but
> >  I have problems at the moment to install it, so I cannot directly compare.
> >
> >  - mh
> >  
>