Re: contour plots and other problems with gnuplot

[Henry F. Mollet]

I believe there are a  number of problems when it comes to graphing in
octave via gnuplot because it's difficult to know what the defaults are and
how to change them.  I usually have to proceed by trial and error because I
cannot figure out what's going on (which may indeed not mean much). Often,
in order to get back to default graphs, I have to quit the terminal and
start from scratch.

When using a script with mesh plots and/or gsplot, it
uses default view  60 rot_x, 30 rot_z, 1 scale, 1 scale_z.
When now using a contour plot command at the octave prompt, the
view changes to 0 rot_x, 0 rot_z, 1 scale, 1 scale_z.
If the script with mesh and/or gsplots is tried again, all plots will be
contour plots because view is not changed back to default. This is actually
not a bad thing because it's much faster to produce a contour plot and a
contour plot may be more easily interpreted than a complicated 3-d graph.

Similarly and disconcerting, I've noticed the following changes when using
gset multiplot and then going back to a single plot:
(Note that "gshow all" produces about 3 pages worth of settings, most of
which I don't understand)

1) Hidden surface is removed (vs. drawn in default).
2) Aspect ratio is 1:1 (vs. not controlled in default).
3) Rectangular grid is drawn in x,y plane at x, y tics (vs. no grid in
default).
5) Changes in "set xrange" and "set yrange" but not sure what's going on.
4) Colors are red and green vs. red only in default but I haven't yet been
able to figure out what this is determined by in the gshow output.
Henry



on 5/2/04 4:29 PM, Victor Munoz at address@hidden wrote:

>> 
>> Does anybody know of a counturc compatible function available under the
>> GPL ?. I would settle for something close which can be easily adapted.
>> 
>> This function should take a matrix and an isoline value as input and
>> return a list of vertexes for the isoline(contour) corresponding to the
>> isovalue.
>> 
>> I have looked in the contour implementation in several GPL'd programs
>> such as epstk and euler, but the do not consider the isoline in it's
>> entirety, rather they plot each little segment of  the line independently.
>> 
> 
> Hello. Maybe PLPlot?
> 
> I've been playing with this problem for a while. I've never felt
> comfortable with the fact that contours are calculated by gnuplot, instead
> of octave calculating them and then passing gnuplot only the lines to draw.
> I usually need to modify contour plots, adding lines, reusing lines in
> certain ways, etc., and that's hard if gnuplot treats them like a type of 3d
> plots (or at least that's how I remember the problem),  and if I don't have
> the data myself available as a matrix in octave for further manipulation.
> 
> For a couple of years I used a contour routine called conrec, which I
> downloaded from http://astronomy.swin.edu.au/~pbourke/projection/conrec/,
> but they also plot little segments. Enough for me, but now I need whole
> lines, so I had to go back to the problem. I chose PLPlot, and currently I'm
> working on extracting the relevant lines of code to build a contour routine
> to use with octave, a la contourc. This only started a couple of days ago, so
> I haven't made much progress, but that's the idea.
> 
> Regards,
> 
>   Victor.
> 
> 
> 
> 
> -------------------------------------------------------------
> Octave is freely available under the terms of the GNU GPL.
> 
> Octave's home on the web:  http://www.octave.org
> How to fund new projects:  http://www.octave.org/funding.html
> Subscription information:  http://www.octave.org/archive.html
> -------------------------------------------------------------



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------