PLplot

I have plplot-5.7.1
There are some data showing time evolution.
I'd like to have scale  of x axis large than y.
I define it via plvpor  (see code below)
The ps driver rotate the picture on -90.
Can I avoid this with plplot tools?
Other drivers, like xwin or png, do job as expected.
Note, if I define the aspect ratio less than 1 (y scale > x) ps driver do job 
nicely but picture is ugly.
    
best regards
Valery

##########
the code for python

def plph ():
    plscol0(1,0,0,0); 
    plsdev("png");plsfnam("intm1.png");
    #plsdev("psttf");plsfnam("intm1.ps");
    plscolbg(255,255,255);
    plinit();
    pladv(0);
    plvpor(0.1, 0.9, 0.3, 0.8);
    plwind(min(time),max(time), min(emfh),max(emfh));
    #plenv(min(time),max(time), min(emfh),max(emfh),0,-1);
    plwid(2);
    pllsty(1);plcol0(14);
    plline(time,emfh)
    pllsty(1);plcol0(1);
    plbox("bnit", 0, 0, "bntv", 0, 0);
    pllab("TIME","E#dz","#gy=0");
    plend()

On 2007-01-26 16:15+0800 Valery Pipin wrote:

> The ps driver rotate the picture on -90.

That's long been the standard for postscript plots in the default
landscape mode.  The only way they will fit on paper is to rotate them so
the shorter y axis corresponds to the shorter width of the page so PLplot
does that rotation by default. Postscript
viewers should know all about portrait versus landscape.  For gv, try the
landscape option to view the default landscape PLplot postscript plots
correctly.

If you actually do want to make a portrait plot (where x is the short axis
aligned with the short edge of the paper) with the postscript devices, use
the -portrait command-line option for your PLplot python script, then use
the gv portrait option to view it.

Alan
__________________________
Alan W. Irwin

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

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); 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
__________________________

Dear Alan,

Thank you for reply and clarifications.
> That's long been the standard for postscript plots in the default
> landscape mode.  The only way they will fit on paper is to rotate them so
> the shorter y axis corresponds to the shorter width of the page so PLplot
> does that rotation by default. Postscript
> viewers should know all about portrait versus landscape.  For gv, try the
> landscape option to view the default landscape PLplot postscript plots
> correctly.
>
> If you actually do want to make a portrait plot (where x is the short axis
> aligned with the short edge of the paper) with the postscript devices, use
> the -portrait command-line option for your PLplot python script, then use
Do you know how to do this within emacs?

regards
Valery

On 2007-01-27 10:00+0800 Valery Pipin wrote:

>> If you actually do want to make a portrait plot (where x is the short axis
>> aligned with the short edge of the paper) with the postscript devices, use
>> the -portrait command-line option for your PLplot python script, then use

> Do you know how to do this within emacs?

No since I am unfamiliar with that operating system.... :-)

Seriously, I have heard you can do any command-line command from within
emacs so I am pretty sure it should be possible to run python scripts with
options from emacs, but somebody else will have to help here who is familiar
with emacs.

Alan
__________________________
Alan W. Irwin

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

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); 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
__________________________

Alan W. Irwin wrote:
> On 2007-01-27 10:00+0800 Valery Pipin wrote:
> >> If you actually do want to make a portrait plot (where x is the short
> >> axis aligned with the short edge of the paper) with the postscript
> >> devices, use the -portrait command-line option for your PLplot python
> >> script, then use
> >
> > Do you know how to do this within emacs?
>
> No since I am unfamiliar with that operating system.... :-)
>
> Seriously, I have heard you can do any command-line command from within
> emacs so I am pretty sure it should be possible to run python scripts with
> options from emacs, but somebody else will have to help here who is
> familiar with emacs.
I can not get it working from command line 
I use "python pli.py -dev psttf -o intm1.ps -portrait".
It ask me about driver nevertheless.
test_python.sh  does not work for me
 [vv@... examples]$ ./test_python.sh
/usr/bin/python: can't open file '/x01': [Errno 2] No such file or directory

Another question,  is how to get \mathcal{E} in psttf driver?

On 2007-01-27 12:47+0800 Valery Pipin wrote:

> I can not get it working from command line
> I use "python pli.py -dev psttf -o intm1.ps -portrait".
> It ask me about driver nevertheless.
> test_python.sh  does not work for me

All the scripts in the test directory need a lot of setting up to work
correctly from the build tree.  Our ctest command does all that, but it is
tough for humans to remember everything to do.  To greatly simplify your
life run "make install" (which copies needed files into rational places in
the install tree), then cd to $prefix/share/plplot-5.7.2/examples/python,
and you should be able to run the ./x?? python scripts without problems from
the command line.  If you follow that model with your own python scripts you
should not have any problems.

Alan
__________________________
Alan W. Irwin

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

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); 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
__________________________

Sorry I didnot ask clearly.
Alan W. Irwin wrote:
> On 2007-01-27 12:47+0800 Valery Pipin wrote:
> > I can not get it working from command line
> > I use "python pli.py -dev psttf -o intm1.ps -portrait".
Is this a correct line?

another question was
how to get \mathcal{E} with pstff driver

regards,
Valery

I'm trying to use the new procedure to build this release.
I have used
"cmake -DCMAKE_INSTALL_PREFIX=/usr/local -DCMAKE_VERBOSE_MAKEFILE=ON  ../plplot-5.7.2 
>& cmake.out"
 
in cmake.out I have
-- WARNING: swig not found. Disabling java bindings
-- WARNING: swig not found. Disabling python bindings
I don't need java. I need python.
[vv@... ~]$ swig
Must specify an input file. Use -help for available options.
[vv@... ~]$ rpm -q swig
swig-1.3.31-alt1

What should I do?

On 2007-01-27 14:49+0800 Valery Pipin wrote:

> I'm trying to use the new procedure to build this release.
> I have used
> "cmake -DCMAKE_INSTALL_PREFIX=/usr/local -DCMAKE_VERBOSE_MAKEFILE=ON ../plplot-5.7.2
>> & cmake.out"

That seems fine (although I have a personal preference for more unique
install prefixes like /usr/local/plplot which makes your plplot install much
easier to remove.)

>
> in cmake.out I have
> -- WARNING: swig not found. Disabling java bindings
> -- WARNING: swig not found. Disabling python bindings
> I don't need java. I need python.
> [vv@... ~]$ swig
> Must specify an input file. Use -help for available options.
> [vv@... ~]$ rpm -q swig
> swig-1.3.31-alt1
>
> What should I do?

The simplest thing to do is to use the cvs version of PLplot.  I have
recently (just after 5.7.2 was released) made a substantial improvement
that makes it much easier for cmake to find swig.  All it needs is
swig in the PATH, and apparently that is what you have so it should work.
It works fine for me, but let me know if you have any problems with it.

Alan
__________________________
Alan W. Irwin

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

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); 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
__________________________

Alan W. Irwin wrote:
> On 2007-01-27 14:49+0800 Valery Pipin wrote:
> > I'm trying to use the new procedure to build this release.
> > I have used
> > "cmake -DCMAKE_INSTALL_PREFIX=/usr/local -DCMAKE_VERBOSE_MAKEFILE=ON
> > ../plplot-5.7.2
> >
> >> & cmake.out"
>
> That seems fine (although I have a personal preference for more unique
> install prefixes like /usr/local/plplot which makes your plplot install
> much easier to remove.)
Does it mean that cmake have no "make uninstall" ?
>
> > in cmake.out I have
> > -- WARNING: swig not found. Disabling java bindings
> > -- WARNING: swig not found. Disabling python bindings
> > I don't need java. I need python.
> > [vv@... ~]$ swig
> > Must specify an input file. Use -help for available options.
> > [vv@... ~]$ rpm -q swig
> > swig-1.3.31-alt1
> >
> > What should I do?
>
> The simplest thing to do is to use the cvs version of PLplot.  I have
I'll try soon  but not today.
Anyway ./configure make & make install works fine.

On 2007-01-27 14:23+0800 Valery Pipin wrote:

> Sorry I didnot ask clearly.
> Alan W. Irwin wrote:
>> On 2007-01-27 12:47+0800 Valery Pipin wrote:
>>> I can not get it working from command line
>>> I use "python pli.py -dev psttf -o intm1.ps -portrait".
> Is this a correct line?

Yes, once PLplot is installed, and you let python know where the PLplot
python modules are installed, that is the correct way to call a python
script with PLplot command-line options.  But first things first; get
the PLplot install working (see previous e-mail) before worrying about this.

>
> another question was
> how to get \mathcal{E} with pstff driver

I can highly recommend the gucharmap application which allows you to look at
all unicode blocks and how your installed system fonts look for those
blocks.  For example, I have the cmsy10 TrueType font (which is a TrueType
representation of the Latex math symbol fonts) installed on my system, and
indeed when I switch to viewing that with gucharmap it looks like the
calligraphic version of characters are in the normal latin alphabet places.
So if you render an upper case E with the cmsy10 TrueType font, it appears
you will get what you want.  First, make sure the cmsy10 TrueType font is
installed on your system.  After that take the following PLplot steps: (a)
specify the TrueType font cmsy10 as one of the 30 fonts that can be
described by an FCI; (b) embed that FCI (Font Characterization Integer) in
your string to change to that font; (c) follow that FCI in the string with
an "E"; and (d) follow that "E" with another FCI to change back to your
original string font. See
http://plplot.sourceforge.net/docbook-manual/plplot-html-5.7.2/characters.html
for details of how to do all these steps.

Alan
__________________________
Alan W. Irwin

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

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); 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
__________________________

>
> Yes, once PLplot is installed, and you let python know where the PLplot
> python modules are installed, that is the correct way to call a python
> script with PLplot command-line options.  But first things first; get
> the PLplot install working (see previous e-mail) before worrying about
> this.
All python scripts from share/examples works fine.
But they  are designed to run interactively. 
Everything is ok in this way.
However they do not understand comand line options.
Ok I'll try it on Ubuntu Feisty and say it is there
>
> > another question was
> > how to get \mathcal{E} with pstff driver
>
> I can highly recommend the gucharmap application which allows you to look
> at all unicode blocks and how your installed system fonts look for those
> blocks.  For example, I have the cmsy10 TrueType font (which is a TrueType
> representation of the Latex math symbol fonts) installed on my system, and
> indeed when I switch to viewing that with gucharmap it looks like the
> calligraphic version of characters are in the normal latin alphabet places.
> So if you render an upper case E with the cmsy10 TrueType font, it appears
> you will get what you want.  First, make sure the cmsy10 TrueType font is
> installed on your system.
they are in latex-ttf package
> After that take the following PLplot steps: (a) 
> specify the TrueType font cmsy10 as one of the 30 fonts that can be
> described by an FCI; (b) embed that FCI (Font Characterization Integer) in
> your string to change to that font; (c) follow that FCI in the string with
> an "E"; and (d) follow that "E" with another FCI to change back to your
> original string font. See
> http://plplot.sourceforge.net/docbook-manual/plplot-html-5.7.2/characters.h
>tml for details of how to do all these steps.
Thanks for explanation!

regards
Valery

On 2007-01-27 20:09+0800 Valery Pipin wrote:

> All python scripts from share/examples works fine.
> But they  are designed to run interactively.
> Everything is ok in this way.
> However they do not understand comand line options.

If the installed python scripts do not understand command-line options then
something is not being done correctly with your install.

Just to give a concrete example, do you get results from (the installed
version) of either of the following commands:

python/x01 -h

or

python python/x01 -h

?

For me, this -h command-line option results in a display of help concerning
all command-line options.  Note, the menu prompting for device should _not_
appear.

Similarly, I get good results from, e.g.,

python/x01 -dev psttfc -o test.ps

or (to return to the original topic)

python/x01 -dev psttfc -o test.ps -portrait

For both these commands, the menu for selecting the device does not appear,
and the command just works with no further interaction since it has everything
it needs from the command-line options.

This is how all the installed examples should work.  For example, after
building the installed examples with the "make" command (of course, this
build is not required for the python examples above), then you should be
able to run

c/x01c -h
c/x01c -dev psttfc -o test.ps
c/x01c -dev psttfc -o test.ps -portrait

etc.  There is something severely wrong with your PLplot install if some/all
of the above does not work properly for you.

Alan
__________________________
Alan W. Irwin

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

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); 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
__________________________