plplot-general — PLplot user discussion group

Re: [Plplot-general] Linking in Linux

[Alan W. I. <Ala...@gm...>]

On 2020-07-22 20:38+0100 Peter Williams wrote:

> Hello Plplot
> I am at present migrating some PlPlot programs from Windows 10 to linux
> (mint 64 bit),
> To facilitate this I have downloaded cmake and Plplot-5.15.0.tar.gz from
> your website.
> Following the instructions from your wiki I built an installation
> consisting of static
> libraries. In order to test the installation  I wrote a very simple program
> PlPlotTest.c:
>
> #include <stdio.h>
> #include "plplot.h"
> int main()
> {
> printf("Plotting example\n");
> plsdev("ps"); plsfnam("TestPlplot.ps");plinit();
> plenv(-1.0 , 1.0 ,-1.0,1.0,1,2);
> pljoin(-0.5,-0.5,0.5,0.5);
> plend();
> printf("Plot finished\n");
> return 1;
> }
> and attempted to compile it with
>
> gcc -c -I /home/peter/plplot/install_directory/include/plplot PlPlotTest.c
> gcc -o PlPlotTest.exe PlPlotTest.o  -L
> /home/peter/plplot/install_directory/lib  -lplplot -lcsirocsa  -lqsastime
> -lm -lX11
>
> An object file PlPlotTest.o was generated with no errors, but the linking
> failed with [undefined pthread library references].

> I am very new to linux. Can you tell me what the linking should be?

Hi Peter:

If you execute "man gcc" and look for the -pthread option it says

"-pthread
  Define additional macros required for using the POSIX threads library.  You should use this option
  consistently for both compilation and linking."

Therefore, I suggest you add the -pthread option to *both* the compile
(gcc -c ....) and link (gcc -o ....)  steps above to see if those
changes solve solve the pthread library linking issues you have
encountered.

Alan

__________________________
Alan W. Irwin

Research affiliation with the Department of Physics and Astronomy,
University of Victoria, Victoria, BC, Canada.

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

[Plplot-general] Linking in Linux

[Peter W. <pet...@gm...>]

Hello Plplot
I am at present migrating some PlPlot programs from Windows 10 to linux
(mint 64 bit),
To facilitate this I have downloaded cmake and Plplot-5.15.0.tar.gz from
your website.
Following the instructions from your wiki I built an installation
consisting of static
libraries. In order to test the installation  I wrote a very simple program
PlPlotTest.c:

#include <stdio.h>
#include "plplot.h"
int main()
{
 printf("Plotting example\n");
 plsdev("ps"); plsfnam("TestPlplot.ps");plinit();
 plenv(-1.0 , 1.0 ,-1.0,1.0,1,2);
 pljoin(-0.5,-0.5,0.5,0.5);
 plend();
 printf("Plot finished\n");
 return 1;
}
and attempted to compile it with

gcc -c -I /home/peter/plplot/install_directory/include/plplot PlPlotTest.c
gcc -o PlPlotTest.exe PlPlotTest.o  -L
/home/peter/plplot/install_directory/lib  -lplplot -lcsirocsa  -lqsastime
-lm -lX11

An object file PlPlotTest.o was generated with no errors, but the linking
failed with the messages:

/usr/bin/ld:
/home/peter/plplot/install_directory/lib/libplplot.a(xwin.c.o): in function
`plD_init_xw':
xwin.c:(.text+0x470): undefined reference to `pthread_mutexattr_init'

/usr/bin/ld: xwin.c:(.text+0x481): undefined reference to
`pthread_mutexattr_settype'
/usr/bin/ld: xwin.c:(.text+0x519): undefined reference to `pthread_create'
/usr/bin/ld:
/home/peter/plplot/install_directory/lib/libplplot.a(xwin.c.o): in function
`plD_tidy_xw':
xwin.c:(.text+0xcd0): undefined reference to `pthread_cancel'
/usr/bin/ld: xwin.c:(.text+0xcec): undefined reference to `pthread_join'
collect2: error: ld returned 1 exit status

I am very new to linux. Can you tell me what the linking should be?
Best Wishes, Peter

-- 
Peter Williams
pet...@gm...
27 Ramsbury Road, St. Albans, AL1 1SN, UK

Re: [Plplot-general] No Color with Ubuntu18.04 Linux

[Alan W. I. <Ala...@gm...>]

On 2020-06-13 15:35-0400 Baumgardner, John R (Mechanical Engineering) wrote:

> Hi Alan,
>
> I attempted to install Plplot-15-5.0 on my Ubuntu 18.04 Linux machine, but 
> cmake could not find where the cairo, qt, tk,etc. libraries were located for 
> some reason.  Rather than attempting to deal with those issues, ....

Hi John:

Well, the usual issue when CMake cannot find libraries is the
"development" versions of the cairo, qt, and tk library packages have not been
installed.  Which should be an easy problem to fix....

> However [when attempting to use 3rd-party PLplot for a later Ubuntu], it would be nice if the xwin option also worked.  
> Currently, when I select that option, I get the error: "*** PLPLOT ERROR, 
> IMMEDIATE EXIT ***  Can't open display  Program aborted."  I am guessing that 
> I am missing something simple in how the X environment is specified.  Might 
> you have any suggestions for resolving that issue?

Actually, I have never had a problem like that with -dev xwin so you
might be up against a packaging error of some kind.  To confirm or
eliminate that possibility, what happens for the version you build
yourself (as above, but with all relevant library development packages
installed)?

If the version you build yourself also has the problem, then I suggest
you try the -debug option which should give more details on what
exactly is going wrong for -dev xwin.

Good luck, and let us know how it goes.

Alan
__________________________
Alan W. Irwin

Research affiliation with the Department of Physics and Astronomy,
University of Victoria, Victoria, BC, Canada.

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Re: [Plplot-general] No Color with Ubuntu18.04 Linux

[Alan W. I. <Ala...@gm...>]

On 2020-06-12 13:44-0400 Baumgardner, John R (Mechanical Engineering) wrote:

> Hi Alan,
>
> I am seeking to utilize the third-party Plplot binary package available on 
> the Ubuntu 18.04 Linux distribution with Fortran.  My application compiles 
> with gfortran and link with all the required libraries with no errors or 
> warnings.  However, the plots that are generated for all the file formats 
> under Cairo and qt and also svg have all the B&W features correct but have no 
> color.  Yet I do get the correct color features for xwin and color 
> PostScript. Might you be able to diagnose what might be awry, based on this 
> brief summary?

Hi John:

I have never experienced such a no-colour symptom with some but not
all of the Plplot device drivers.  Therefore, I am forwarding your
question to the Plplot list to see if anyone there has ever had
similar experience on Ubuntu or any other platform.

To check if there is some peculiar Ubuntu packaging error, what
are your results for a PLplot version that you build yourself?
[Our wiki](https://sourceforge.net/p/plplot/wiki/Home/) should have all
the information you require to make such a build.

However, for this build please do the following:

# Download Plplot-5.15.0 as per wiki instructions.
# Configure the PLplot build with cmake as per wiki instructions.
cmake ...
# Build the C "00" example.
make x00c
# Build the svg device
make svg
# Create a plot with that device
examples/c/x00c -dev svg -o test.svg

Then assuming that test.svg result has no colour for however you view
it on your system, please send that test.svg result to this list to
see whether it has colour when we display it with our own SVG viewers.

I have just gone through the above process myself (for the latest git
version of PLplot rather than 5.15.0, but I don't think that will make
a difference), and I have attached the test.svg file that I generate
here with the above commands.  The correct colour (red lines and
letters on black background) shows up both for my favorite SVG viewer
(pqiv) as well as firefox (which can also be used as a SVG viewer by
simply typing the complete pathname of the file in the URL box).  When
you display that file on your system with your favorite SVG viewer, is
the correct colour displayed there?

Alan
__________________________
Alan W. Irwin

Research affiliation with the Department of Physics and Astronomy,
University of Victoria, Victoria, BC, Canada.

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Re: [Plplot-general] Log scale for frequency plot:

[Raphael R. <raf...@bl...>]

Thank you Alan,
I'll try that asap!

rph


Le 18.05.20 à 18:59, Alan W. Irwin a écrit :
> On 2020-05-05 21:45+0200 Raphael Raccuia wrote:
>
>> Hi list,
>> Any chance to get a custom x axis log scale?
>
> Hi Raphael:
>
> Thanks for your interest in PLplot, and the answer to your question is
> "yes".
>
> The best way to figure out how to do anything with PLplot
> is to consult the code for one of our standard examples.
>
> For example:
>
> # Search for case-insensitive "custom" in our C example source code.
> software@merlin> grep -i custom examples/c/*.c
> examples/c/x02c.c:// Use 100 custom colors in addition to base 16
> examples/c/x02c.c:// Load default cmap0 colors into our custom set
> examples/c/x19c.c:// A custom axis labeling function for longitudes 
> and latitudes.
> examples/c/x19c.c:    // Setup a custom latitude and longitude-based 
> scaling function.
> examples/c/x19c.c:    plscol0( 8, 100, 100, 100 ); //dark grey for 
> custom polygons - generally crags
> examples/c/x33c.c:    "Shade colorbars with custom labels",
>
> So example 19 looks promising for your needs.  Here is software for 
> that example for our various languages:
>
> software@merlin> ls -l examples/*/*19*
> -rw-r--r-- 1 software software 15621 Jan 27 22:26 
> examples/ada/xstandard19a.adb
> -rw-r--r-- 1 software software 14616 Jan 27 22:26 
> examples/ada/xtraditional19a.adb
> -rw-r--r-- 1 software software 12104 Jan 27 22:26 examples/c++/x19.cc
> -rw-r--r-- 1 software software  9911 Jan 27 22:26 examples/c/x19c.c
> -rw-r--r-- 1 software software  9420 Jan 27 22:26 examples/d/x19d.d
> -rw-r--r-- 1 software software 16424 Jan 27 22:26 
> examples/fortran/x19f.f90
> -rw-r--r-- 1 software software 12440 Jan 27 22:26 examples/java/x19.java
> -rw-r--r-- 1 software software  8549 Jan 27 22:26 examples/lua/x19.lua
> -rw-r--r-- 1 software software  9405 Jan 27 22:26 examples/ocaml/x19.ml
> -rw-r--r-- 1 software software  2154 Jan 27 22:26 examples/octave/p19.m
> -rw-r--r-- 1 software software  8532 Jan 27 22:26 examples/octave/x19c.m
> -rwxr-xr-x 1 software software  1158 Jan 27 22:26 examples/python/x19*
> -rw-r--r-- 1 software software  9732 Jan 27 22:26 examples/python/x19.py
> -rwxr-xr-x 1 software software   462 Jan 27 22:26 examples/tcl/x19.in*
> -rw-r--r-- 1 software software  9425 Jan 27 22:26 examples/tcl/x19.tcl
>
> Pick the language of interest to you, and go from there.  Or if you 
> prefer
> a GUI then consult <http://plplot.org/examples.php?demo=19>.
>
> Alan
> __________________________
> Alan W. Irwin
>
> Research affiliation with the Department of Physics and Astronomy,
> University of Victoria, Victoria, BC, Canada.
>
> Programming affiliations with the FreeEOS equation-of-state
> implementation for stellar interiors (freeeos.sf.net); the Time
> Ephemerides project (timeephem.sf.net); PLplot scientific plotting
> software package (plplot.org); the libLASi project
> (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
> and the Linux Brochure Project (lbproject.sf.net).
> __________________________
>
> Linux-powered Science
> __________________________

Re: [Plplot-general] Log scale for frequency plot:

[Alan W. I. <Ala...@gm...>]

On 2020-05-05 21:45+0200 Raphael Raccuia wrote:

> Hi list,
> Any chance to get a custom x axis log scale?

Hi Raphael:

Thanks for your interest in PLplot, and the answer to your question is
"yes".

The best way to figure out how to do anything with PLplot
is to consult the code for one of our standard examples.

For example:

# Search for case-insensitive "custom" in our C example source code.
software@merlin> grep -i custom examples/c/*.c
examples/c/x02c.c:// Use 100 custom colors in addition to base 16
examples/c/x02c.c:// Load default cmap0 colors into our custom set
examples/c/x19c.c:// A custom axis labeling function for longitudes and latitudes.
examples/c/x19c.c:    // Setup a custom latitude and longitude-based scaling function.
examples/c/x19c.c:    plscol0( 8, 100, 100, 100 ); //dark grey for custom polygons - generally crags
examples/c/x33c.c:    "Shade colorbars with custom labels",

So example 19 looks promising for your needs.  Here is software for that example for our various languages:

software@merlin> ls -l examples/*/*19*
-rw-r--r-- 1 software software 15621 Jan 27 22:26 examples/ada/xstandard19a.adb
-rw-r--r-- 1 software software 14616 Jan 27 22:26 examples/ada/xtraditional19a.adb
-rw-r--r-- 1 software software 12104 Jan 27 22:26 examples/c++/x19.cc
-rw-r--r-- 1 software software  9911 Jan 27 22:26 examples/c/x19c.c
-rw-r--r-- 1 software software  9420 Jan 27 22:26 examples/d/x19d.d
-rw-r--r-- 1 software software 16424 Jan 27 22:26 examples/fortran/x19f.f90
-rw-r--r-- 1 software software 12440 Jan 27 22:26 examples/java/x19.java
-rw-r--r-- 1 software software  8549 Jan 27 22:26 examples/lua/x19.lua
-rw-r--r-- 1 software software  9405 Jan 27 22:26 examples/ocaml/x19.ml
-rw-r--r-- 1 software software  2154 Jan 27 22:26 examples/octave/p19.m
-rw-r--r-- 1 software software  8532 Jan 27 22:26 examples/octave/x19c.m
-rwxr-xr-x 1 software software  1158 Jan 27 22:26 examples/python/x19*
-rw-r--r-- 1 software software  9732 Jan 27 22:26 examples/python/x19.py
-rwxr-xr-x 1 software software   462 Jan 27 22:26 examples/tcl/x19.in*
-rw-r--r-- 1 software software  9425 Jan 27 22:26 examples/tcl/x19.tcl

Pick the language of interest to you, and go from there.  Or if you prefer
a GUI then consult <http://plplot.org/examples.php?demo=19>.

Alan
__________________________
Alan W. Irwin

Research affiliation with the Department of Physics and Astronomy,
University of Victoria, Victoria, BC, Canada.

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

[Plplot-general] Log scale for frequency plot

[Raphael R. <raf...@bl...>]

Hi list,
Any chance to get a custom x axis log scale? I want to draw a log grid 
with ticks and labels at for example:
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
6300
8000
10000
12500
16000

(which correspond to frequencies of an equalizer)

Thank you!
rph-r

Re: [Plplot-general] remove margins on multiple plot scenario

[Alan W. I. <Ala...@gm...>]

On 2020-02-20 18:07+0100 Xavier Cardil wrote:

> Thank you Alan
>
> I have successfully integrated your solution in to my program

You are welcome, and I am glad I was able to help you find a good
solution to the "remove margins" issue you raised.

Best wishes,

Alan
__________________________
Alan W. Irwin

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Re: [Plplot-general] remove margins on multiple plot scenario

[Xavier C. <car...@gm...>]

Thank you Alan

I have successfully integrated your solution in to my program

At the end, in order to do it I had to build something similar to plssub so
it played out with plstick

As stacking was just fine, I just split the pipe in to plssub + plenv for
stacking and my grid builder / indexer + plstick

Thank you again for the in-depth explanation, the code example and the link
to the documentation, which I recognize I should've checked before
impulsively start coding :)

Have fun!

Xavier

On Sat, Feb 15, 2020 at 7:27 PM Alan W. Irwin <Ala...@gm...>
wrote:

> On 2020-02-14 15:01+0100 Xavier Cardil wrote:
>
> > Hi Alan:
> > Thank you for sharing your code !
>
> You are welcome.
>
> >
> > I have been trying to integrate this in to my current python program
> > without luck, as it's already pretty complex, so what I am doing now
> > My program creates custom "subplot" objects and passes them through a
> > "pipe" that first generates the window, then organizes the plots and
> > finally draws them
> > One of the critical parts of the procedure is using plenv, which uses
> > itself pladv and sets a basic viewport, as per the plplot documentation
> > As it's hard to implement this without an example, I am trying to
> implement
> > your solution on a regular plplot example first, but:
> >
> > I have some difficulty understanding your code, so I would really
> > appreciate if you can clear some doubts:
>
> I think you already understand the plstick documentation (see my
> response to your documentation questions below).  And actually, the
> code is pretty clear as well.  Aside from some scaling of characters
> and symbols (which I don't completely follow myself, but it works
> reasonably well), it is just a wrapper for plvpor but with a different
> argument list that is convenient to use when you are grouping plots close
> together on a single page.
>
> Please review
> <http://plplot.org/docbook-manual/plplot-html-5.15.0/viewport_window.html>
> to make sure you understand the distinctions between plotting surface,
> (sub-)page, and viewport, and the plwind and plbox calls that
> establish the window corresponding to the viewport and annotate the
> viewport.
>
> For example, you could make a 4 plots jammed together for a given page
> by
>
> plinit()
> pladv(0)
>
> # Upper left viewport on page
> plvpor(0., 0.5, 0.5, 1.0)
> plwind(...)
> ...
> plbox(...)
> ...
>
> # Lower left viewport on page
> plvpor(0., 0.5, 0.0, 0.5)
> plwind(...)
> ...
> plbox(...)
> ...
>
> # Upper right viewport on page
> plvpor(0.5, 1.0, 0.5, 1.0)
> plwind(...)
> ...
> plbox(...)
> ...
>
> # Lower right viewport on page
> plvpor(0.5, 1.0, 0.0, 0.5)
> plwind(...)
> ...
> plbox(...)
> ...
>
> But you can successively replace those plvpor calls by
>
> plstick(2,2,1,1, 0., 0., 0., 0.)
> plstick(2,2,2,1, 0., 0., 0., 0.)
> plstick(2,2,1,2, 0., 0., 0., 0.)
> plstick(2,2,2,2, 0., 0., 0., 0.)
>
> to achieve exactly the same results (aside from the character and symbol
> scaling that
> goes on inside plstick).  Note, those first 2 arguments specify the size
> of the
> matrix of potential viewports, and the 3rd and 4 arguments specify the
> matrix
> element corresponding to the single viewport established by the
> plvpor call within plstick.
>
> The fundamental advantages of plstick over plvpor is those third and
> fourth arguments makes it easy to keep track of which part of the
> viewport matrix you are attempting to use, and you don't have to worry
> about keeping track of all the divisions and offsets to specify a
> particular element of the matrix.  For example, to specify the (2, 3)
> viewport matrix element within a 3x7 matrix you just need to call
>
> plstick(3,7,2,3,....)
>
> Or you could call plvpor directly instead, but that case there would be a
> lot of
> arithmetic involved to generate the correct limits corresponding to
> that matrix element.
>
> Note plenv internally calls pladv, plvpor, plwind, and plbox with some
> simplifying assumptions so it should not be used for any situation
> that is complex in the slightest
> such as attempting to jam plot viewports (and corresponding
> windows) together on a single (sub-)page.  So instead of plenv you
> must (as in the example above) call pladv once per (sub-)page followed
> by groups of plvpor (or plstick), plwind, and plbox calls that jam the
> viewports (and corresponding windows) together.
>
> If you want to take a look at a more complex but real-world example of
> using plstick, see
> <
> https://sourceforge.net/p/freeeos/freeeos/ci/master/tree/www/eos_papers/coulomb/coulomb_adjust.py
> >.
> This is the updated version of the Python script that generated the
> plot I attached last time.  If you look at that plot, there are two
> columns and 3 rows for the viewport matrix that is used, but a
> complication is there is a gap between the two columns to leave room
> for the annotation of the right-most column.  But specifying such a
> gap (and the space above and below each column) is handled with ease using
> the plstick spacing parameters.
>
> >
> >
> >   - # Sets up a viewport such that xnum X ynum of them fit onto a # page.
> >      - I assume this sets the amount of plots that will fit in to a page
> X
> >      / Y distributed ?
>
> Yes.
>
>
> >   - Make the active viewport the one at column currentx and row #
> >   currenty, starting at the top left.
> >      - This sets where plots will start being drawn, starting at the top
> >      left ?
>
> Yes.
>
> So I have confirmed you understand the documentation of plstick, and you
> have access to general documentation of the relationships between
> page, viewport, and window and two examples to work with.  So I hope
> this information completely solves the issue for you.
>
> Alan
> __________________________
> Alan W. Irwin
>
> Programming affiliations with the FreeEOS equation-of-state
> implementation for stellar interiors (freeeos.sf.net); the Time
> Ephemerides project (timeephem.sf.net); PLplot scientific plotting
> software package (plplot.org); the libLASi project
> (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
> and the Linux Brochure Project (lbproject.sf.net).
> __________________________
>
> Linux-powered Science
> __________________________
>

Re: [Plplot-general] remove margins on multiple plot scenario

[Alan W. I. <Ala...@gm...>]

On 2020-02-14 15:01+0100 Xavier Cardil wrote:

> Hi Alan:
> Thank you for sharing your code !

You are welcome.

>
> I have been trying to integrate this in to my current python program
> without luck, as it's already pretty complex, so what I am doing now
> My program creates custom "subplot" objects and passes them through a
> "pipe" that first generates the window, then organizes the plots and
> finally draws them
> One of the critical parts of the procedure is using plenv, which uses
> itself pladv and sets a basic viewport, as per the plplot documentation
> As it's hard to implement this without an example, I am trying to implement
> your solution on a regular plplot example first, but:
>
> I have some difficulty understanding your code, so I would really
> appreciate if you can clear some doubts:

I think you already understand the plstick documentation (see my
response to your documentation questions below).  And actually, the
code is pretty clear as well.  Aside from some scaling of characters
and symbols (which I don't completely follow myself, but it works
reasonably well), it is just a wrapper for plvpor but with a different
argument list that is convenient to use when you are grouping plots close
together on a single page.

Please review
<http://plplot.org/docbook-manual/plplot-html-5.15.0/viewport_window.html>
to make sure you understand the distinctions between plotting surface,
(sub-)page, and viewport, and the plwind and plbox calls that
establish the window corresponding to the viewport and annotate the viewport.

For example, you could make a 4 plots jammed together for a given page
by

plinit()
pladv(0)

# Upper left viewport on page
plvpor(0., 0.5, 0.5, 1.0)
plwind(...)
...
plbox(...)
...

# Lower left viewport on page
plvpor(0., 0.5, 0.0, 0.5)
plwind(...)
...
plbox(...)
...

# Upper right viewport on page
plvpor(0.5, 1.0, 0.5, 1.0)
plwind(...)
...
plbox(...)
...

# Lower right viewport on page
plvpor(0.5, 1.0, 0.0, 0.5)
plwind(...)
...
plbox(...)
...

But you can successively replace those plvpor calls by

plstick(2,2,1,1, 0., 0., 0., 0.)
plstick(2,2,2,1, 0., 0., 0., 0.)
plstick(2,2,1,2, 0., 0., 0., 0.)
plstick(2,2,2,2, 0., 0., 0., 0.)

to achieve exactly the same results (aside from the character and symbol scaling that
goes on inside plstick).  Note, those first 2 arguments specify the size of the
matrix of potential viewports, and the 3rd and 4 arguments specify the matrix
element corresponding to the single viewport established by the
plvpor call within plstick.

The fundamental advantages of plstick over plvpor is those third and
fourth arguments makes it easy to keep track of which part of the
viewport matrix you are attempting to use, and you don't have to worry
about keeping track of all the divisions and offsets to specify a
particular element of the matrix.  For example, to specify the (2, 3)
viewport matrix element within a 3x7 matrix you just need to call

plstick(3,7,2,3,....)

Or you could call plvpor directly instead, but that case there would be a lot of
arithmetic involved to generate the correct limits corresponding to
that matrix element.

Note plenv internally calls pladv, plvpor, plwind, and plbox with some
simplifying assumptions so it should not be used for any situation
that is complex in the slightest
such as attempting to jam plot viewports (and corresponding
windows) together on a single (sub-)page.  So instead of plenv you
must (as in the example above) call pladv once per (sub-)page followed
by groups of plvpor (or plstick), plwind, and plbox calls that jam the
viewports (and corresponding windows) together.

If you want to take a look at a more complex but real-world example of
using plstick, see
<https://sourceforge.net/p/freeeos/freeeos/ci/master/tree/www/eos_papers/coulomb/coulomb_adjust.py>.
This is the updated version of the Python script that generated the
plot I attached last time.  If you look at that plot, there are two
columns and 3 rows for the viewport matrix that is used, but a
complication is there is a gap between the two columns to leave room
for the annotation of the right-most column.  But specifying such a
gap (and the space above and below each column) is handled with ease using
the plstick spacing parameters.

>
>
>   - # Sets up a viewport such that xnum X ynum of them fit onto a # page.
>      - I assume this sets the amount of plots that will fit in to a page X
>      / Y distributed ?

Yes.


>   - Make the active viewport the one at column currentx and row #
>   currenty, starting at the top left.
>      - This sets where plots will start being drawn, starting at the top
>      left ?

Yes.

So I have confirmed you understand the documentation of plstick, and you
have access to general documentation of the relationships between
page, viewport, and window and two examples to work with.  So I hope
this information completely solves the issue for you.

Alan
__________________________
Alan W. Irwin

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Re: [Plplot-general] remove margins on multiple plot scenario

[Xavier C. <car...@gm...>]

Hi Alan:
Thank you for sharing your code !

I have been trying to integrate this in to my current python program
without luck, as it's already pretty complex, so what I am doing now
My program creates custom "subplot" objects and passes them through a
"pipe" that first generates the window, then organizes the plots and
finally draws them
One of the critical parts of the procedure is using plenv, which uses
itself pladv and sets a basic viewport, as per the plplot documentation
As it's hard to implement this without an example, I am trying to implement
your solution on a regular plplot example first, but:

I have some difficulty understanding your code, so I would really
appreciate if you can clear some doubts:


   - # Sets up a viewport such that xnum X ynum of them fit onto a # page.
      - I assume this sets the amount of plots that will fit in to a page X
      / Y distributed ?
   - Make the active viewport the one at column currentx and row #
   currenty, starting at the top left.
      - This sets where plots will start being drawn, starting at the top
      left ?
   - The "space" parameters (top, # right, bot, and left) specify how much
   space is to be left open at # the relevant side for annotations etc.
      - This is clear


I believe it would also be great if you could share a full example with
this implementation, if possible.

Thank you so much !

Xavier





On Sat, Feb 8, 2020 at 8:05 PM Alan W. Irwin <Ala...@gm...>
wrote:

> On 2020-02-07 13:14+0100 Xavier Cardil wrote:
>
> > I was wondering how could I remove the white space between plots when
> using
> > a grid-like subplot configuration. When I stack plots vertically, there
> is
> > a small margin that is negligible for our use, but not when on a grid
> > scenario ( see picture )
> >
> > [image: plplot.png]
> >
> > If you guys could give me a fast tip on how to control the margins
> between
> > plots, or just to remove them completely I'd appreciate it
>
> Hi Xavier:
>
> It is interesting you brought up this topic since I have recently
> encountered it myself.  Some of my older research plots used the
> "yplot" wrapper for PLplot which dealt with this issue by defining the
> "plstick" functionality, but yplot is moribund so when recently
> attempting to replicate those old research plots I copied the
> essentials of plstick to Python so I could use plstick from that
> language.  I attach a figure showing the type of results you can get
> with this Python version of plstick, and if you like those I suggest
> you port the Python version of plstick (which I have also attached) to
> whatever language you like to use to generate your plots.
>
> Alan
> __________________________
> Alan W. Irwin
>
> Programming affiliations with the FreeEOS equation-of-state
> implementation for stellar interiors (freeeos.sf.net); the Time
> Ephemerides project (timeephem.sf.net); PLplot scientific plotting
> software package (plplot.org); the libLASi project
> (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
> and the Linux Brochure Project (lbproject.sf.net).
> __________________________
>
> Linux-powered Science
> __________________________

Re: [Plplot-general] remove margins on multiple plot scenario

[Alan W. I. <Ala...@gm...>]

On 2020-02-07 13:14+0100 Xavier Cardil wrote:

> I was wondering how could I remove the white space between plots when using
> a grid-like subplot configuration. When I stack plots vertically, there is
> a small margin that is negligible for our use, but not when on a grid
> scenario ( see picture )
>
> [image: plplot.png]
>
> If you guys could give me a fast tip on how to control the margins between
> plots, or just to remove them completely I'd appreciate it

Hi Xavier:

It is interesting you brought up this topic since I have recently
encountered it myself.  Some of my older research plots used the
"yplot" wrapper for PLplot which dealt with this issue by defining the
"plstick" functionality, but yplot is moribund so when recently
attempting to replicate those old research plots I copied the
essentials of plstick to Python so I could use plstick from that
language.  I attach a figure showing the type of results you can get
with this Python version of plstick, and if you like those I suggest
you port the Python version of plstick (which I have also attached) to
whatever language you like to use to generate your plots.

Alan
__________________________
Alan W. Irwin

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

[Plplot-general] remove margins on multiple plot scenario

[Xavier C. <car...@gm...>]

I was wondering how could I remove the white space between plots when using
a grid-like subplot configuration. When I stack plots vertically, there is
a small margin that is negligible for our use, but not when on a grid
scenario ( see picture )

[image: plplot.png]

If you guys could give me a fast tip on how to control the margins between
plots, or just to remove them completely I'd appreciate it

Thanks  !

Xavier

Re: [Plplot-general] drawing boxes, keeps on painting ?

[Hazen B. <hba...@ma...>]

Maybe the problem is that you are using "vecsize" instead of "5" in the
call to p.line()?

p.line(vecsize,px,py);
versus
p.line(5,px,py);

I'm guessing that vecsize is not always equal to 5. However it would
help if you show all the code, and also the working Python version for
comparison.

best,
-Hazen

On 1/21/20 1:38 PM, Xavier Cardil wrote:
> Hi, any comment on this ? I really can't find the reason lines seem to
> get off viewport box randomly, even after several tries refactoring this
> function ... any help would be appreciated !
> 
> On Thu, Jan 16, 2020 at 4:04 PM Xavier Cardil <car...@gm...
> <mailto:car...@gm...>> wrote:
> 
>     Hi, I am trying to draw boxes following example 32
>     <http://plplot.sourceforge.net/examples.php?demo=32&lbind=C++>. I
>     successfully implemented this on Python, but there is some weird
>     problem when trying to do it on C++:
> 
>     Sometimes, boxes are drawn correctly :
> 
> 
>     20200116_160142.png
> 
>     But sometimes, ( seems random ) this happens :
> 
> 
>     20200116_155042.png
> 
>     Is this something I am missing ? I can't figure out why is this
>     happening?
> 
>     Below is my code, which is really simple. Thanks
> 
> 
>     plstream p;
> 
>     const auto hmax = max_element(begin(vec.high), end(vec.high));
>     const auto lmin = min_element(begin(vec.low), end(vec.low));
> 
>     vector<int> ivec(vecsize);
> 
>     generate(ivec.begin(), ivec.end(), [steps = 0]() mutable { return
>     steps++; });
>     const char *device = "xwin";
>     p.start(device, 1, 1);
>     p.ssub(0, 1);
>     p.env(0, vecsize, *lmin, *hmax, 0, 0);
> 
>     PLFLT px[5], py[5];
>     PLFLT wx[5], wy[5];
>     PLFLT spacing;
>     PLFLT xmid;
> 
>     spacing = .4;// in x axis units
> 
>     for (int i = 0; i < vecsize; i++) {
>     px[0]=ivec[i]+spacing /2.;
>     py[0]=vec.open[i];
>     px[1]=ivec[i]+spacing /2.;
>     py[1]=vec.close[i];
>     px[2]=ivec[i]+1.-spacing /2.;
>     py[2]=vec.close[i];
>     px[3]=ivec[i]+1.-spacing /2.;
>     py[3]=vec.open[i];
>     px[4]=ivec[i]+spacing /2.;
>     py[4]=vec.open[i];
> 
>     p.col0(1);
>     p.line(vecsize,px,py);
> 
> 
> 
> _______________________________________________
> Plplot-general mailing list
> Plp...@li...
> https://lists.sourceforge.net/lists/listinfo/plplot-general
> 

Re: [Plplot-general] drawing boxes, keeps on painting ?

[Xavier C. <car...@gm...>]

Hi, any comment on this ? I really can't find the reason lines seem to get
off viewport box randomly, even after several tries refactoring this
function ... any help would be appreciated !

On Thu, Jan 16, 2020 at 4:04 PM Xavier Cardil <car...@gm...>
wrote:

> Hi, I am trying to draw boxes following example 32
> <http://plplot.sourceforge.net/examples.php?demo=32&lbind=C++>. I
> successfully implemented this on Python, but there is some weird problem
> when trying to do it on C++:
>
> Sometimes, boxes are drawn correctly :
>
>
> [image: 20200116_160142.png]
>
> But sometimes, ( seems random ) this happens :
>
>
> [image: 20200116_155042.png]
>
> Is this something I am missing ? I can't figure out why is this happening?
>
> Below is my code, which is really simple. Thanks
>
>
> plstream p;
>
> const auto hmax = max_element(begin(vec.high), end(vec.high));
> const auto lmin = min_element(begin(vec.low), end(vec.low));
>
> vector<int> ivec(vecsize);
>
> generate(ivec.begin(), ivec.end(), [steps = 0]() mutable { return steps++;
> });
> const char *device = "xwin";
> p.start(device, 1, 1);
> p.ssub(0, 1);
> p.env(0, vecsize, *lmin, *hmax, 0, 0);
>
> PLFLT px[5], py[5];
> PLFLT wx[5], wy[5];
> PLFLT spacing;
> PLFLT xmid;
>
> spacing = .4; // in x axis units
>
> for (int i = 0; i < vecsize; i++) {
> px[0] = ivec[i] + spacing / 2.;
> py[0] = vec.open[i];
> px[1] = ivec[i] + spacing / 2.;
> py[1] = vec.close[i];
> px[2] = ivec[i] + 1. - spacing / 2.;
> py[2] = vec.close[i];
> px[3] = ivec[i] + 1. - spacing / 2.;
> py[3] = vec.open[i];
> px[4] = ivec[i] + spacing / 2.;
> py[4] = vec.open[i];
>
> p.col0(1);
> p.line(vecsize, px, py);
>

[Plplot-general] drawing boxes, keeps on painting ?

[Xavier C. <car...@gm...>]

Hi, I am trying to draw boxes following example 32
<http://plplot.sourceforge.net/examples.php?demo=32&lbind=C++>. I
successfully implemented this on Python, but there is some weird problem
when trying to do it on C++:

Sometimes, boxes are drawn correctly :


[image: 20200116_160142.png]

But sometimes, ( seems random ) this happens :


[image: 20200116_155042.png]

Is this something I am missing ? I can't figure out why is this happening?

Below is my code, which is really simple. Thanks


plstream p;

const auto hmax = max_element(begin(vec.high), end(vec.high));
const auto lmin = min_element(begin(vec.low), end(vec.low));

vector<int> ivec(vecsize);

generate(ivec.begin(), ivec.end(), [steps = 0]() mutable { return steps++;
});
const char *device = "xwin";
p.start(device, 1, 1);
p.ssub(0, 1);
p.env(0, vecsize, *lmin, *hmax, 0, 0);

PLFLT px[5], py[5];
PLFLT wx[5], wy[5];
PLFLT spacing;
PLFLT xmid;

spacing = .4; // in x axis units

for (int i = 0; i < vecsize; i++) {
px[0] = ivec[i] + spacing / 2.;
py[0] = vec.open[i];
px[1] = ivec[i] + spacing / 2.;
py[1] = vec.close[i];
px[2] = ivec[i] + 1. - spacing / 2.;
py[2] = vec.close[i];
px[3] = ivec[i] + 1. - spacing / 2.;
py[3] = vec.open[i];
px[4] = ivec[i] + spacing / 2.;
py[4] = vec.open[i];

p.col0(1);
p.line(vecsize, px, py);

Re: [Plplot-general] plstream instantiation [C++]

[Xavier C. <car...@gm...>]

It was a matter of how I built Plpot, it's fixed now, thanks

On Thu, Jan 9, 2020 at 3:53 PM Xavier Cardil <car...@gm...>
wrote:

> Following this document :
> https://www.absoft.com/wp-content/uploads/2015/08/plplot-5.3.1.pdf
>
> The C++ programmer can write: plstream p( ... ); p.env( ... ); p.line( ...
>> );
>
>
> I have "started" to port my Python program to C++, and trying to
> instantiate plstream and, as example, call start() and give it some
> parameters:
>
> const char *x = "xwin";
>
> PLINT once = 1;
>
> PLINT twice = 1;
>
> cout << x << endl;
>
> plstream p;
>
> p.start(x, once, twice);
>
> But I get this at compile:
>
> [build] CMakeFiles/autoplot++.dir/main.cpp.o: In function `main':
> [build] /home/xavier/autoplot_plus/main.cpp:521: undefined reference to
> `plstream::plstream()'
> [build] /home/xavier/autoplot_plus/main.cpp:521: undefined reference to
> `plstream::~plstream()'
> [build] collect2: error: ld returned 1 exit status
> [build] make[2]: *** [autoplot++] Error 1
> [build] make[1]: *** [CMakeFiles/autoplot++.dir/all] Error 2
> [build] make: *** [all] Error 2
>
> I'm not sure what's missing here, I'd appreciate if you guys can give me a
> hint
>
> Thanks
>
> Xavier
>
>

[Plplot-general] plstream instantiation [C++]

[Xavier C. <car...@gm...>]

Following this document :
https://www.absoft.com/wp-content/uploads/2015/08/plplot-5.3.1.pdf

The C++ programmer can write: plstream p( ... ); p.env( ... ); p.line( ...
> );


I have "started" to port my Python program to C++, and trying to
instantiate plstream and, as example, call start() and give it some
parameters:

const char *x = "xwin";

PLINT once = 1;

PLINT twice = 1;

cout << x << endl;

plstream p;

p.start(x, once, twice);

But I get this at compile:

[build] CMakeFiles/autoplot++.dir/main.cpp.o: In function `main':
[build] /home/xavier/autoplot_plus/main.cpp:521: undefined reference to
`plstream::plstream()'
[build] /home/xavier/autoplot_plus/main.cpp:521: undefined reference to
`plstream::~plstream()'
[build] collect2: error: ld returned 1 exit status
[build] make[2]: *** [autoplot++] Error 1
[build] make[1]: *** [CMakeFiles/autoplot++.dir/all] Error 2
[build] make: *** [all] Error 2

I'm not sure what's missing here, I'd appreciate if you guys can give me a
hint

Thanks

Xavier

Re: [Plplot-general] Help building plplot

[Arjen M. <Arj...@de...>]

Hi Walt,

I had a look at the files you attached and compared them to typical output from CMake on my machine. The errors you got about crt_externs.h and a division by zero are actually harmless. Or should be. They are part of the probing procedure – determine what flavours of header files are available and other such system-specific stuff. If the test returns an error, then the corresponding flag is set to false (see plplot_config.h.in for instance). The Fortran test should be revised, as the gfortran compiler performs the division, not the program. But other than that there should be no problem.

Does CMake produce such output as:

-- Configuring done
-- Generating done
-- Build files have been written to: …

If so, then the next step is to run make explicitly.

If not, could you send me the output of CMake (cmake -G “Unix Makefiles ../source-directory > cmake.out 2>&1) and the files in the build directory?

Also – VERY IMPORTANT – start with a completely empty build directory. Left-overs from a previous build tend to confuse CMake/make.

Regards,

Arjen

From: Walt Brainerd <wal...@gm...>
Sent: 03 January 2020 17:50
To: plplot_general <plp...@li...>
Subject: Re: [Plplot-general] Help building plplot

Thanks again Arjen.

Stupidity:   I simply omitted the -G option.
Unfortunately, using it ("Unix Makefiles") produced the same result.

Ignorance: Using the cygwin "make" produced the following (I assume crt_externs.h is some cygwin
file, but I don't know how/where to get it without installing all of cygwin (days worth of downloading), but
will do that if nothing else works.


On Thu, Jan 2, 2020 at 3:09 PM Walt Brainerd <wal...@gm...<mailto:wal...@gm...>> wrote:
Thanks Arjen for the suggestions. I always learn something from you.
However, I still can't get where I want.

With the plplot that comes with Msys2, execution hangs awaiting input.
If I type "1", it produces output on the screen. Also -dev works (Arjen).

Trying to build plplot with msys produces similar (or worse) results.

Trying build with cygwin, cmake produces the following lines multiple times.

$ "/cygdrive/c/Program Files/Cmake/bin/cmake.exe" -DCMAKE_INSTALL_PREFIX=install -DENABLE_cxx=OFF -DENABLE_cxx=OFF -DENABLE_qt=OFF ../plplot.git >& cmake.out

-- Configuring done
CMake Error in bindings/fortran/CMakeLists.txt:
  Evaluation file to be written multiple times with different content.  This
  is generally caused by the content evaluating the configuration type,
  language, or location of object files:

   C:/walt/Software/Plplot/BUILD/bindings/fortran/plplotfortran_ifort.def.sh<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Fplplotfortran_ifort.def.sh&data=02%7C01%7C%7Cfd1b4cf9bfd14e4b1ebd08d7906d0de4%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637136670380210211&sdata=ZwOoEbpCsnLJx4xTlW8OhZjq38kpG%2FlLg%2FsgJ3SRhTQ%3D&reserved=0>

I am using the latest cmake (3.16.2) and the latest plplot. cygwin does not
have the latest Fortran compiler, but that shouldn't be a problem, should it?
cmake.out is attached.

Any help would be appreciated. OW, I will just use the flaky msys plplot.

Thanks.



--
Walt Brainerd


--
Walt Brainerd
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] Help building plplot

[Walt B. <wal...@gm...>]

Thanks again Arjen.

Stupidity:   I simply omitted the -G option.
Unfortunately, using it ("Unix Makefiles") produced the same result.

Ignorance: Using the cygwin "make" produced the following (I assume
crt_externs.h is some cygwin
file, but I don't know how/where to get it without installing all of cygwin
(days worth of downloading), but
will do that if nothing else works.


On Thu, Jan 2, 2020 at 3:09 PM Walt Brainerd <wal...@gm...>
wrote:

> Thanks Arjen for the suggestions. I always learn something from you.
> However, I still can't get where I want.
>
> With the plplot that comes with Msys2, execution hangs awaiting input.
> If I type "1", it produces output on the screen. Also -dev works (Arjen).
>
> Trying to build plplot with msys produces similar (or worse) results.
>
> Trying build with cygwin, cmake produces the following lines multiple
> times.
>
> $ "/cygdrive/c/Program Files/Cmake/bin/cmake.exe"
> -DCMAKE_INSTALL_PREFIX=install -DENABLE_cxx=OFF -DENABLE_cxx=OFF
> -DENABLE_qt=OFF ../plplot.git >& cmake.out
>
> -- Configuring done
> CMake Error in bindings/fortran/CMakeLists.txt:
>   Evaluation file to be written multiple times with different content.
> This
>   is generally caused by the content evaluating the configuration type,
>   language, or location of object files:
>
>    C:/walt/Software/Plplot/BUILD/bindings/fortran/
> plplotfortran_ifort.def.sh
>
> I am using the latest cmake (3.16.2) and the latest plplot. cygwin does not
> have the latest Fortran compiler, but that shouldn't be a problem, should
> it?
> cmake.out is attached.
>
> Any help would be appreciated. OW, I will just use the flaky msys plplot.
>
> Thanks.
>
>
>
> --
> Walt Brainerd
>


-- 
Walt Brainerd

Re: [Plplot-general] Help building plplot

[Arjen M. <Arj...@de...>]

Hi Walt,

Are you using the Cygwin version of CMake? I am not sure what the differences are, but I know there are some (possibly just the makefile generator …). I see you are using the Visual Studio generator – I am not sure that is a good idea in combination with Cygwin. The reason for that is that for some platforms (Cygwin and MinGW-w64/MSYS2) on Windows a bash shell is used and for “bare” Windows it is not. CMake tries to hide these platform issues as much as possible, but that does not mean (unfortunately) that you can use any combination 😉.

With a fairly recent version of gfortran you should be able to build PLplot. Some further comments below.

Regards,

Arjen

From: Walt Brainerd <wal...@gm...>
Sent: 02 January 2020 23:09
To: plplot_general <plp...@li...>
Subject: [Plplot-general] Help building plplot

Thanks Arjen for the suggestions. I always learn something from you.
However, I still can't get where I want.

>>AM: you make me blush 😊.

With the plplot that comes with Msys2, execution hangs awaiting input.
If I type "1", it produces output on the screen. Also -dev works (Arjen).

Trying to build plplot with msys produces similar (or worse) results.

>>AM: nastiness with buffering of output to the screen, sigh.

Trying build with cygwin, cmake produces the following lines multiple times.

$ "/cygdrive/c/Program Files/Cmake/bin/cmake.exe" -DCMAKE_INSTALL_PREFIX=install -DENABLE_cxx=OFF -DENABLE_cxx=OFF -DENABLE_qt=OFF ../plplot.git >& cmake.out

-- Configuring done
CMake Error in bindings/fortran/CMakeLists.txt:
  Evaluation file to be written multiple times with different content.  This
  is generally caused by the content evaluating the configuration type,
  language, or location of object files:

   C:/walt/Software/Plplot/BUILD/bindings/fortran/plplotfortran_ifort.def.sh<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Fplplotfortran_ifort.def.sh&data=02%7C01%7C%7C15270ecebb7e4a6cba7908d78fd07610%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637135997820643311&sdata=fI0JAP1SzRf2stUr6gfPjsjGGo2gnwbXcYZnNMcZ2nk%3D&reserved=0>

I am using the latest cmake (3.16.2) and the latest plplot. cygwin does not
have the latest Fortran compiler, but that shouldn't be a problem, should it?
cmake.out is attached.

Any help would be appreciated. OW, I will just use the flaky msys plplot.

Thanks.

>>AM: I strongly suspect this is because of the mixture of platforms. Try:

  *   CMake for Cygwin (according to Cygwin’s setup: the current version is 3.14.5-1)
  *   Use the generator options -G "Unix Makefiles" (this is the appropriate generator for Cygwin)
  *   Cygwin has gcc/gfortran 7.4.0-1, that is certainly recent enough



--
Walt Brainerd
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

[Plplot-general] Help building plplot

[Walt B. <wal...@gm...>]

Thanks Arjen for the suggestions. I always learn something from you.
However, I still can't get where I want.

With the plplot that comes with Msys2, execution hangs awaiting input.
If I type "1", it produces output on the screen. Also -dev works (Arjen).

Trying to build plplot with msys produces similar (or worse) results.

Trying build with cygwin, cmake produces the following lines multiple times.

$ "/cygdrive/c/Program Files/Cmake/bin/cmake.exe"
-DCMAKE_INSTALL_PREFIX=install -DENABLE_cxx=OFF -DENABLE_cxx=OFF
-DENABLE_qt=OFF ../plplot.git >& cmake.out

-- Configuring done
CMake Error in bindings/fortran/CMakeLists.txt:
  Evaluation file to be written multiple times with different content.  This
  is generally caused by the content evaluating the configuration type,
  language, or location of object files:

   C:/walt/Software/Plplot/BUILD/bindings/fortran/plplotfortran_ifort.def.sh

I am using the latest cmake (3.16.2) and the latest plplot. cygwin does not
have the latest Fortran compiler, but that shouldn't be a problem, should
it?
cmake.out is attached.

Any help would be appreciated. OW, I will just use the flaky msys plplot.

Thanks.



-- 
Walt Brainerd

Re: [Plplot-general] Plplot in Msys2

[Arjen M. <Arj...@de...>]

Hi Walt,

I have seen such behaviour before – with some environments output to the console is either buffered or dropped. The command-line options, like “./a -dev wingcc” should work though.

If the program stops immediately when run in a Windows console, then there is a good chance that it stops because it could not find one or more of the DLLs it depends on. If you run it from a file manager like Windows Explorer you may get a message box showing the name of the missing DLL. My tool of choice to find out which is “dependency walker” – http://www.dependencywalker.com/ or the newer rewrite at https://github.com/lucasg/Dependencies.

Regards,

Arjen

From: Walt Brainerd <wal...@gm...>
Sent: 24 December 2019 20:43
To: Alan W. Irwin <Ala...@gm...>
Cc: plplot_general <plp...@li...>
Subject: Re: [Plplot-general] Plplot in Msys2

Things have changed a bit with my tinkering.

Pls ignore the statements about the examples--all bogus.
I do have 5.15.0 in my Msys2.

When I compile in an Msys window, ./a just hangs. If I type a "1",
the graph is generated on the screen. The display options usually
do not show up at all.

When I compile with exactly the same compiler in a Windows
command window, it executes and terminates immediately with
no output.

Apparently running an Msys terminal sets something that helps,,
but it is still not right.

If this suggests anything to anybody, please let me know.
In the meantime, I will try to build plplot from source.

I have no confidence that it is the fault of Msys, given my
lack of expertise about all this, so am reluctant to submit
a bug. But maybe I will if I can my own build to work.

Thanks.

On Mon, Dec 23, 2019 at 8:25 PM Alan W. Irwin <Ala...@gm...<mailto:Ala...@gm...>> wrote:
On 2019-12-23 14:40-0700 Walt Brainerd wrote:

[...]
> I see the Fortran examples [for official MSYS2 package version] are quite different, but I did use the latest.
> Why were they rewritten?
[...]

Hi Walt:

The Fortran examples were historically rewritten a long time ago but
not any time recently.  Therefore, I am virtually positive you are
seeing an old version of the PLplot fortran examples because you have
accessed an old official MSYS2 package version of PLplot.

What happens if you do a general update of your MSYS2 platform
following the [MSYS2 instructions for doing
that](https://github.com/msys2/msys2/wiki/MSYS2-installation<https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fmsys2%2Fmsys2%2Fwiki%2FMSYS2-installation&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815636572&sdata=p5Hk9W4J5Um9OqEORnV1DgDSN4u8SNQenCeWVQfGuH8%3D&reserved=0>)?  My bet
is you will then have access to the latest official packaged version
of PLplot which is our latest release (5.15.0).

By the way, we are normally quite happy with the MSYS2 platform for
PLplot.  For example, as part of the 5.15.0 pre-release testing, Arjen
has had good Fortran (and much other) success on MSYS2 with a PLplot
version that he built himself.  However, as far as I know he has never
tested the official MSYS2 PLplot package there which has been packaged
by the principal packager for MSYS2 (alexpux) who is familiar with
MSYS2 but not PLplot.  So if it turns out that official MSYS2 packaged
version of PLplot-5.15.0 does not work, you should prepare a bug
report for MSYS2 and revert to using a 5.15.0 version that you built
yourself on MSYS2 just to follow closely what Arjen did for his test.
And with any luck at all, you should be able to replicate Arjen's
Fortran success with that method.

Alan
__________________________
Alan W. Irwin

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Ffreeeos.sf.net&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815636572&sdata=%2FCMEqUG4IkWXqu7s%2Bn3rp8IfSVo5f%2BuEt96axOVAGFM%3D&reserved=0>); the Time
Ephemerides project (timeephem.sf.net<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Ftimeephem.sf.net&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815646563&sdata=sxtjLdA%2FO1mXtp5wQd3ezLJEV71Khsjq9Sk5sbeZ2og%3D&reserved=0>); PLplot scientific plotting
software package (plplot.org<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Fplplot.org&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815646563&sdata=s3v3YIs34rLGq7GCYnveCdmqocgQfycNuBjwFhANTJc%3D&reserved=0>); the libLASi project
(unifont.org/lasi<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Funifont.org%2Flasi&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815646563&sdata=BbXaULKZ0U0L6XdOrPA9XySQlWVynHMJimVQ2xbT4Kc%3D&reserved=0>); the Loads of Linux Links project (loll.sf.net<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Floll.sf.net&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815656556&sdata=4MD%2FUjk%2BL2jROUv05PyU%2B9%2FLdjYTbi2zr7rGxBpwbKA%3D&reserved=0>);
and the Linux Brochure Project (lbproject.sf.net<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Flbproject.sf.net&data=02%7C01%7C%7C831500c32ab84bfb131908d788a97af0%7C15f3fe0ed7124981bc7cfe949af215bb%7C0%7C0%7C637128133815656556&sdata=EysoSTvxSmGekuNMEH%2BztYZXazvdLMprgdyG2m7wdIc%3D&reserved=0>).
__________________________

Linux-powered Science
__________________________


--
Walt Brainerd
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] Plplot in Msys2

[Walt B. <wal...@gm...>]

Things have changed a bit with my tinkering.

Pls ignore the statements about the examples--all bogus.
I do have 5.15.0 in my Msys2.

When I compile in an Msys window, ./a just hangs. If I type a "1",
the graph is generated on the screen. The display options usually
do not show up at all.

When I compile with exactly the same compiler in a Windows
command window, it executes and terminates immediately with
no output.

Apparently running an Msys terminal sets something that helps,,
but it is still not right.

If this suggests anything to anybody, please let me know.
In the meantime, I will try to build plplot from source.

I have no confidence that it is the fault of Msys, given my
lack of expertise about all this, so am reluctant to submit
a bug. But maybe I will if I can my own build to work.

Thanks.

On Mon, Dec 23, 2019 at 8:25 PM Alan W. Irwin <Ala...@gm...>
wrote:

> On 2019-12-23 14:40-0700 Walt Brainerd wrote:
>
> [...]
> > I see the Fortran examples [for official MSYS2 package version] are
> quite different, but I did use the latest.
> > Why were they rewritten?
> [...]
>
> Hi Walt:
>
> The Fortran examples were historically rewritten a long time ago but
> not any time recently.  Therefore, I am virtually positive you are
> seeing an old version of the PLplot fortran examples because you have
> accessed an old official MSYS2 package version of PLplot.
>
> What happens if you do a general update of your MSYS2 platform
> following the [MSYS2 instructions for doing
> that](https://github.com/msys2/msys2/wiki/MSYS2-installation)?  My bet
> is you will then have access to the latest official packaged version
> of PLplot which is our latest release (5.15.0).
>
> By the way, we are normally quite happy with the MSYS2 platform for
> PLplot.  For example, as part of the 5.15.0 pre-release testing, Arjen
> has had good Fortran (and much other) success on MSYS2 with a PLplot
> version that he built himself.  However, as far as I know he has never
> tested the official MSYS2 PLplot package there which has been packaged
> by the principal packager for MSYS2 (alexpux) who is familiar with
> MSYS2 but not PLplot.  So if it turns out that official MSYS2 packaged
> version of PLplot-5.15.0 does not work, you should prepare a bug
> report for MSYS2 and revert to using a 5.15.0 version that you built
> yourself on MSYS2 just to follow closely what Arjen did for his test.
> And with any luck at all, you should be able to replicate Arjen's
> Fortran success with that method.
>
> Alan
> __________________________
> Alan W. Irwin
>
> Programming affiliations with the FreeEOS equation-of-state
> implementation for stellar interiors (freeeos.sf.net); the Time
> Ephemerides project (timeephem.sf.net); PLplot scientific plotting
> software package (plplot.org); the libLASi project
> (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
> and the Linux Brochure Project (lbproject.sf.net).
> __________________________
>
> Linux-powered Science
> __________________________
>


-- 
Walt Brainerd

Re: [Plplot-general] problem building

[Alan W. I. <Ala...@gm...>]

On 2019-12-24 13:51+0100 Xavier Cardil wrote:

> I am now porting a program from Python to C++ that uses Plplot
>
> Getting this error when running Cmake in my project :
>
> [cmake] CMake Error at lib/plplot.git/CMakeLists.txt:60
> (cmake_minimum_required):
> [cmake]   CMake 3.13.2 or higher is required.  You are running version
> 3.10.2
>
> But my system Cmake version is cmake version 3.16.2
>
> This is my Makelists.txt:
>
> cmake_minimum_required(VERSION 3.0.0)
> project(autoplot++ VERSION 0.1.0)
>
> include(CTest)
> enable_testing()
>
> add_executable(autoplot++ main.cpp)
>
> add_subdirectory(lib/yaml-cpp)
>
> add_subdirectory(lib/plplot.git)
>
> include_directories(
> lib/yaml-cpp/include,
> lib/plplot.git/bindings/c++
> )
>
> target_link_libraries(
> autoplot++
> PRIVATE
> yaml-cpp
> )
>
> target_link_libraries(
> autoplot++
> PRIVATE
> plplot
> )
>
> set(CPACK_PROJECT_NAME ${PROJECT_NAME})
> set(CPACK_PROJECT_VERSION ${PROJECT_VERSION})
> include(CPack)
>

Hi Xavier:

If PLplot is complaining that you are using CMake-3.10.2,
then that is very likely the case.  So if running Linux
or some other Unix, try

which cmake

from the commmand line to see where you are picking up that old version.

Also, to diagnose the problem more directly (and my experience is that
using old versions of cmake is a bad idea so that is really a problem) I
would change to

cmake_minimum_required(VERSION 3.13.2)

in your own project.

Beyond that, I have never tried building plplot as a sub-project of
another project like you are attempting to do.  I am pretty sure that
method "should" work, but I am also sure there are some PLplot details
that would have to be adjusted upstream by us to make sure it really
does work.  (For example, there is a distinction between
CMAKE_SOURCE_DIR and PROJECT_SOURCE_DIR, and similarly for the
*BINARY* equivalents of those variables which we have not been careful
about.)

So for now I would advise avoiding that method, and instead I suggest
you configure, build, and install PLplot independently.

For such an independent PLplot install, an example of how you access
that install from a different CMake project is given in
examples/CMakeLists.txt for the case where CORE_BUILD is not true
(i.e., this is an example project where the installed source code for
the examples is being built against the installed PLplot libraries).

In other words, your own CMake project should follow what is done in
that file starting at "# MAINTENANCE 2019-02-03"

Good luck, let us know how it goes, and best wishes for the season.

Alan
__________________________
Alan W. Irwin

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.org); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________