Re: [Plplot-devel] Wish list

[Alan W. I. <ai...@us...>]

On 2003-12-31 17:15+0100 Michel Peyrard wrote:

>
>  After Rafael's message about a wish list for further developments of
> plplot, I expressed interest for a program to display images. In fact I
> noticed that it exists in C. It is plimage. But it is not part of the
> fortran library. Would it be possible to include it in fortran too?

Possibly.  But IMO plimage is still a work in progress.  For example, it
cannot handle colour images.  So I have used that limitation as an excuse
not to work on it and instead concentrate on other aspects of improving and
extending the fortran interface to PLplot.  But if somebody else wants to
work on a fortran interface to plimage, they are more than welcome.  Some
information about the relatively simple steps you have to take is given in
bindings/f77/README.f77API.  Essentially you just follow a pattern.

> plshade is not a replacement because it is more complex as it seems to do
> contouring.

You can turn that off which is done in example 15.  Furthermore, I would
advocate almost always turning off contouring (or marking the edge of the
filled area) for plshade since it is rather clumsy when it does do that.  If
you just do one shade the edges are drawn nicely, but subsequent shades can
partially overlay the contour edges due to inevitable rounding errors in the
exact position of the edge of the area being filled by plshade.  So
essentially I have given up on the plshade-style contouring.

plshades (a front end to plshade and plcont) does contouring a different way
by optionally calling plcont after the several calls to plshade (with _its_
edges turned off) to fill the various areas.  Look at plshades results
(example 16) where you will notice some plots have contouring and others not
at user option. I am just now implementing plshades in fortran and will soon
be able to reproduce example 16 from that language.  Will that functionality
satisfy your needs?

> Moreover the possibility to display only one part of the array
> (i.e. have arguments for the dimensions and others for indices of the part
> to plot) would be useful.

The way plshade is implemented in fortran makes it difficult to do this.
(The assumption is made deep in the API that the defined index ranges are
identical to the absolute dimensions.) However, it is possible with some
effort to change the implementation (and also change the fortran API which
will require an additional "leading-dimension" parameter).

But for now, I am leaving the fortran interface to plshade alone and instead
concentrating on the new fortran interface to plshades where one of the
parameters will be the leading dimension (assumed identical) for the 2D
arrays. In other words, it will be set up like the fortran interface to
plot3dc, plmeshc, etc., see current fortran example 8 where the leading
dimension is larger than the defined leading dimension of the 2D area.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the PLplot scientific plotting software
package (plplot.org), the Yorick front-end to PLplot (yplot.sf.net), the
Loads of Linux Links project (loll.sf.net), and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________