gnuplot-beta — Developers and beta testers of new gnuplot code

Re: Replot in epslatex terminal

[<br...@ph...>]

Gaël Varoquaux wrote:
>     Replot in epslatex terminal acts exactly like in postscript terminal
> : it creates a new page of the epsfile. 

It must not do that.  EPS files by their very definition cannot
have multiple pages.

 > But this page never gets
> displayed by latex. 

It shouldn't --- it's not allowed to be there.

 > Wouldn't it be a more consistent option to have
> "replot" act like in the x11, or the png terminal ?

It cannot act like the X11 terminal.  It has to act like (I hope) png 
does: ignore all further plot commands after the first one until the 
output is closed and (another one) re-opened.

Re: Open questions for inclusion into the FAQ?

[<br...@ph...>]

Johannes Nendwich wrote:

> 1) is it possible to define similar to defining strings (filnam =
> "file1.dat"), similar to defining numerical variables (n = 3) ?

There's one too many "similar to" in that sentence, but yes, this is 
possible.  So far only in the development version though.

> 2) is it possible to use different columns from different files in one plot:
>    plot "file1.dat" u 3 : "file2.dat" u 2

No.  For that, use external file proceessing tools, e.g.

	plot "<paste file1.dat file2.dat" u 3:5

> 3) is it possible to create multi-page postscript output?

Yes.  Just leave terminal, output and multiplot alone, don't use the 
'eps' option, and sequence of plots will end up each on a page of its own.

Re: Open questions for inclusion into the FAQ?

[Ethan A M. <merritt@u.washington.edu>]

On Monday 20 June 2005 05:52 pm, Johannes Nendwich wrote:
>
> 1) is it possible to define similar to defining strings (filnam =
> "file1.dat"), similar to defining numerical variables (n = 3) ?

Yes, but only in version 4.1

> 2) is it possible to use different columns from different files in one
> plot: plot "file1.dat" u 3 : "file2.dat" u 2

No.  
However, depending on the file formats it may be possible to
merge the files into one stream as you go:
	NCOLS = 3
	plot "join file1.dat file2.dat" using 3:(NCOLS+2)

> 3) is it possible to create multi-page postscript output?

Of course. Every plot gets a new page. Surely you tried this?

-- 
Ethan A Merritt
Biomolecular Structure Center
University of Washington, Seattle 98195-7742

Re: help exit

[Petr M. <mi...@ph...>]

>>> gnuplot> help gnuplot
>>> TOPIC: gnuplot-defined variables

Good idea. Will it list all options (those "?" lines in gnuplot.doc) for the 
given topic?

> Whether a help for "gnuplot" is needed is even questionable.

gnuplot.doc many times cites `gnuplot`

> but occassionally the user will by chance choose some characters that 
> match a key.

Exactly, it's good to know what I'm actually reading. The GUI hypertext 
help readers show this, but not the default command line help browser.

---
PM

Re: help exit

[Petr M. <mi...@ph...>]

I propose this text:


  The commands `exit` and `quit`, as well as the END-OF-FILE character
  (usually  Ctrl-D) terminate input from the current input stream: terminal
  session,  pipe, and file input (pipe).

  If input streams are nested (inherited `load` scripts), then reading will
  continue in the parent stream. When the top level stream is closed, the
  program itself will exit.

  See "help batch/interactive" for more details.

  Each of these commands will clear the output device (as does the `clear`
  command) before exiting.


BUT: Is the last point right??? Definitely not for -persist option!

---
PM

Re: [ 1027032 ] Connect gnuplot_x11 to exterior application window

[Daniel J S. <dan...@ie...>]

Daniel J Sebald wrote:
> Petr Mikulik wrote:
> 
>>>
>>> How do you mean?
>>
>>
>>
>> To have another tcl script which demonstrates also the mousing 
>> functionality.
> 
> 
> Yes, that would be nice.  I'll search the web to see if there is some 
> way of turning off the moust via tcl.

Sorry Petr, I just cannot seem to get tcl to relinquish use of ButtonPress and 
ButtonRelease.  The X window and tcl literature is either too voluminous or 
arcane and internet searches lead me nowhere.  I'm not even sure it is possible. 
  I've tried things like

bind .plt.g <ButtonPress>
bind .plt.g <ButtonRelease>

event delete {on all the events that "event info" says are active}

Nothing.

However, I HAVE VERIFIED the mouse option by creating an X window via C and 
clearing ButtonPress and ButtonRelease.  gnuplot is then able to utilize the 
mouse in said window.

So, the patch works on my system, but unfortunately tcl can't demostrate the 
mouse portion of it.

I've upload a patch that has a few extra comments in it.

Dan

Re: help exit

[Daniel J S. <dan...@ie...>]

Ethan Merritt wrote:
 > On Wednesday 22 June 2005 01:54 pm, Daniel J Sebald wrote:
 >
 >>Speaking of "help", attached is a short patch for displaying the topic at the
 >>very beginning of help, e.g.,
 >>
 >>gnuplot> help gnuplot
 >>TOPIC: gnuplot-defined variables
 >>  The variable `pi` is defined to be pi, see
 >>        print pi
 >
 >
 >
 > Don't you think a more reasonable fix would be to add a line
 > ?gnuplot
 > at the start of some more appropraite help section?

Yes and no.  Whether a help for "gnuplot" is needed is even questionable.  But
the problem is accounting for all the possible things a user could type
expecting to see help for.  In most cases gnuplot would say "Sorry, no help",
but occassionally the user will by chance choose some characters that match a
key.  Sure, correcting by filling a hole in gnuplot.doc will work, but that's a
potentially never ending game, isn't it?

It's not that critical.

Dan

Re: help exit

[Ethan M. <merritt@u.washington.edu>]

On Wednesday 22 June 2005 01:54 pm, Daniel J Sebald wrote:
> 
> Speaking of "help", attached is a short patch for displaying the topic at the 
> very beginning of help, e.g.,
> 
> gnuplot> help gnuplot
> TOPIC: gnuplot-defined variables
>   The variable `pi` is defined to be pi, see
>         print pi


Don't you think a more reasonable fix would be to add a line
?gnuplot
at the start of some more appropraite help section?


-- 
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Re: help exit

[Daniel J S. <dan...@ie...>]

Ethan Merritt wrote:
> On Tuesday 21 June 2005 02:35 pm, Petr Mikulik wrote:
> 
>>It seems the 'help exit' does not say anything that it exits gnuplot when in 
>>interactive?

Speaking of "help", attached is a short patch for displaying the topic at the 
very beginning of help, e.g.,

gnuplot> help gnuplot
TOPIC: gnuplot-defined variables
  The variable `pi` is defined to be pi, see
        print pi
[...]

I accidentally typed "help gnuplot" and then found it disorienting that what 
returned was actually help about variables.

Dan

Replot in epslatex terminal

[V. <gae...@no...>]

    Replot in epslatex terminal acts exactly like in postscript terminal
: it creates a new page of the epsfile. But this page never gets
displayed by latex. Wouldn't it be a more consistent option to have
"replot" act like in the x11, or the png terminal ?

    Ga=EBl

Re: Key layout patch

[Daniel J S. <dan...@ie...>]

Petr Mikulik wrote:
>> I think the key-cleanup
> 
> 
> Yes, that's interesting. Please update it for current cvs.
> There, I propose to change
>     ... outside | below | <position>}
> to
>     outside | below | at <position>}

Alright, good point.  I'm right in the process of updating this patch with CVS 
changes since last year.

> I see not two but one patch. Please cite patch numbers.

OK, my memory isn't so great either.  I see I combined that into a single patch 
last fall.  I'll drop a note when a new patch is ready.

> 
>> PS:  Is it maybe time to start thinking about a new release?  
>> Developers have been diligent about not letting CVS accumulate bugs; 
>> this is the time of year when most people will have free time; and 
>> after only 14 months the release process should still be fairly fresh 
>> in people's minds.
> 
> 
> Myself, I will have more time in autumn.

That's fine, I think ramping up over a few month time frame is a good idea.

Dan

Re: Key layout patch

[Petr M. <mi...@ph...>]

> I think the key-cleanup

Yes, that's interesting. Please update it for current cvs.
There, I propose to change
 	... outside | below | <position>}
to
 	outside | below | at <position>}

as giving raw numbers is always confusing.


> and enhancement patch(es) is a good one.

I see not two but one patch. Please cite patch numbers.

> PS:  Is it maybe time to start thinking about a new release?  Developers have 
> been diligent about not letting CVS accumulate bugs; this is the time of year 
> when most people will have free time; and after only 14 months the release 
> process should still be fairly fresh in people's minds.

Myself, I will have more time in autumn.

Some ideas for future:

- We should clean patch, bugs and requests section...

- I always apply the "raise" patch, I think we had some proposals via email 
how to change it (new terminal API "term->interactive(...)" to send commands 
to interactive terminals, like "close", "raise", "hide mouse menu", "don't 
use q hotkey to quit" etc, is it is currently done non-consistently over 
terminals.

- I would like to see "set xtics in|out scale ..." to separate them from the 
default "set ticscale ...; set tics in|out".

- Recently I have noticed that it would be convenient to be able do
 	set termoptions close
to be able to close the window under x11/windows/pm et al. terminals (later 
raise, etc.). Can somebody implement this for X11?

---
PM

Re: help exit

[Robert H. <en...@no...>]

On Tue, 2005-06-21 at 23:35 +0200, Petr Mikulik wrote:
>   When running gnuplot in interactive mode, the commands `exit` and `quit`
>   quit your gnuplot session.
> 
>   When reading a command file (script), the commands `exit` and `quit` and
>   the END-OF-FILE character will exit this file immediately and `load` the
>   next one.
> 
>   See "help batch/interactive" for more details.
> 
>   Each of these commands will clear the output device (as does the `clear`
>   command) before exiting.

Isn't ^D considered to be END-OF-FILE for interactive input? It
certainly works, so I don't see the distinction between interactive and
command file.

Rob

-- 
Robert Hart <en...@no...>
University of Nottingham


This message has been checked for viruses but the contents of an attachment
may still contain software viruses, which could damage your computer system:
you are advised to perform your own checks. Email communications with the
University of Nottingham may be monitored as permitted by UK legislation.

Re: help exit

[Ethan M. <merritt@u.washington.edu>]

On Tuesday 21 June 2005 02:35 pm, Petr Mikulik wrote:
> It seems the 'help exit' does not say anything that it exits gnuplot when in 
> interactive?

There is nothing special about interactive sessions. If you would like to
clarify the help text, I suggest something like

   The commands `exit` and `quit` terminate input from the current input
   stream. This is true for terminal, pipe, and file input. If input streams
   are nested, then input will continue to be accepted from the parent
   stream.  When the top level stream is closed, the program itself will exit.



To demonstrate that "interactive" is not relevant to this, try the following:

gnuplot
plot sin(x)
load '/dev/tty'
plot cos(x)
load '/dev/tty'
plot x
exit
history
exit
hisory
exit

You will find that it takes three `exits` to get back to the shell from
this nested interactive session.



 
> What about this text:
> 
>   When running gnuplot in interactive mode, the commands `exit` and `quit`
>   quit your gnuplot session.
> 
>   When reading a command file (script), the commands `exit` and `quit` and
>   the END-OF-FILE character will exit this file immediately and `load` the
>   next one.
> 
>   See "help batch/interactive" for more details.
> 
>   Each of these commands will clear the output device (as does the `clear`
>   command) before exiting.
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> gnuplot-beta mailing list
> gnu...@li...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-beta
> 

-- 
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

help exit

[Petr M. <mi...@ph...>]

It seems the 'help exit' does not say anything that it exits gnuplot when in 
interactive?

What about this text:

  When running gnuplot in interactive mode, the commands `exit` and `quit`
  quit your gnuplot session.

  When reading a command file (script), the commands `exit` and `quit` and
  the END-OF-FILE character will exit this file immediately and `load` the
  next one.

  See "help batch/interactive" for more details.

  Each of these commands will clear the output device (as does the `clear`
  command) before exiting.

Re: linetype palette

[Petr M. <mi...@ph...>]

> that makes the scansforward, scansbackward and scansautomatic options 
> unnecessary.

No. These are important to draw 'pm3d map' from overlapping 3D data (e.g. in 
repetitive overlapping experimental mapping of some property over a sample 
surface).

---
PM

Re: linetype palette

[Petr M. <mi...@ph...>]

> Btw.: how should I proceed to get the depth patch into gnuplot?

I've put some comments into cvs. After having a nice demo of how it works, 
so that we can test it, it will go to cvs after...

---
PM

Re: wxwidgets terminal on the road

[<tim...@en...>]

Daniel J Sebald wrote:

> Timoth=E9e Lecomte wrote:
>
>> Hi !
>>
>> After a long week to dig into C++, wxwidgets, the C code of gnuplot,=20
>> threads, etc. , I am pleased to present a first screenshot of the=20
>> wxwidgets terminal in action :
>>
>> http://tipote.free.fr/wxt.png
>
>
> Looks good Timoth=E9e.
>
> I just wanted to draw your attention to a couple, perhaps unrelated,=20
> items.
>
> First, I just updated a patch that allows attaching an outside=20
> x-window to gnuplot's x-window driver.  (It's patch item #1027032.)  A=20
> tcl demo is included so you can compare that and your methodology.  I=20
> think they are different approaches even though in the long run=20
> similar widgets are being used.
>
> Second, I recall another terminal that works with widgets in some=20
> way.  I think it is called "tkcanvas".  I tried it a while ago and it=20
> created sort of "objectified" plots that could be manipulated in a=20
> third-party program.  (If I'm remembering correctly.)  It seemed very=20
> useful in a way, but didn't appear to be a full featured terminal.
>
> Dan

I think there's a problem with last patches (20/06 and 21/06)... I get=20
errors when opening the .patch (after gunzip) with an editor or with the=20
patch commad ("only garbage ..."). But previous ones seem to work.

I've seen a terminal in maxima based on tcl/tk. It used a small window=20
to give scale and other options to the user... It was not really=20
pleasant to the eyes, but the idea seemed interesting.

I will try your patch as soon as I can open it !

Timoth=E9e

Key layout patch

[Daniel J S. <dan...@ie...>]

Now would be a good time for me to update some of the patches on S.F. if any of 
the developers are interested.

I think the key-cleanup and enhancement patch(es) is a good one.  I recall there 
being two patches for that.  The first patch was only cleanup, adding no new 
functionality but making the current layout options more robust.  What was good 
about this patch, if I remember correctly, is that it took a step toward making 
plot layout for 2D and 3D similar.

The second patch of this pair was one that added a few more options for laying 
out the key so that it can be placed basically anywhere inside or outside the 
plot borders.

Let me know if you are interested in this one.  Also, should it be kept as two 
different patches?  Or should it be combined into one mondo-patch.

Dan

PS:  Is it maybe time to start thinking about a new release?  Developers have 
been diligent about not letting CVS accumulate bugs; this is the time of year 
when most people will have free time; and after only 14 months the release 
process should still be fairly fresh in people's minds.

Re: wxwidgets terminal on the road

[Daniel J S. <dan...@ie...>]

Timoth=E9e Lecomte wrote:
> Hi !
>=20
> After a long week to dig into C++, wxwidgets, the C code of gnuplot,=20
> threads, etc. , I am pleased to present a first screenshot of the=20
> wxwidgets terminal in action :
>=20
> http://tipote.free.fr/wxt.png

Looks good Timoth=E9e.

I just wanted to draw your attention to a couple, perhaps unrelated, item=
s.

First, I just updated a patch that allows attaching an outside x-window t=
o=20
gnuplot's x-window driver.  (It's patch item #1027032.)  A tcl demo is in=
cluded=20
so you can compare that and your methodology.  I think they are different=
=20
approaches even though in the long run similar widgets are being used.

Second, I recall another terminal that works with widgets in some way.  I=
 think=20
it is called "tkcanvas".  I tried it a while ago and it created sort of=20
"objectified" plots that could be manipulated in a third-party program.  =
(If I'm=20
remembering correctly.)  It seemed very useful in a way, but didn't appea=
r to be=20
a full featured terminal.

Dan

Re: wxwidgets terminal

[Daniel J S. <dan...@ie...>]

Daniel J Sebald wrote:
> Ethan Merritt wrote:
> 
>> On Tuesday 14 June 2005 03:09 pm, mi...@ph... wrote:
>>
>>> If it was under X11, have you used the new feature of an x11 handle from
>>> sourceforge patches? This patch waits for confirmation that it works.
> 
> 
> I've updated things on my system.  The demo works as originally designed 
> except for one problem.  The buttons and list items need to be pressed 
> several times to cause the actions of starting a new demo or clearing a 
> demo.  It seems to be a random number of times, but usually two or three 
> times.

Ay, ay, ay... adding a newline character at the front of the gnuplot command 
string right after the ctrl-c appears to have fixed the problem.  E.g., 
("gnuplot" is a pipe command defined in the tcl file)

       gnuplot "\nset term x11; reset; load \"$file\""

So, I've updated the patch on S.F. and things should be ready to go.  The demo 
is run with

command> gpdemos.tcl &

Dan

Open questions for inclusion into the FAQ?

[Johannes N. <nen...@as...>]

hello,

sorry, these are only questions, without answers.

1) is it possible to define similar to defining strings (filnam =
"file1.dat"), similar to defining numerical variables (n = 3) ?

2) is it possible to use different columns from different files in one plot:
   plot "file1.dat" u 3 : "file2.dat" u 2

3) is it possible to create multi-page postscript output?


regards,

john

wxwidgets terminal on the road

[<tim...@en...>]

Hi !

After a long week to dig into C++, wxwidgets, the C code of gnuplot,=20
threads, etc. , I am pleased to present a first screenshot of the=20
wxwidgets terminal in action :

http://tipote.free.fr/wxt.png

Today, the terminal has basic functionnality, as you can see : it can=20
draw lines, text, justify it. But an important work was needed to=20
achieve it :

* mix C and C++, with proper declarations of functions, and with the=20
help of Makefiles
* launch the terminal in a new thread, because every gui is based on an=20
event loop, which doesn't terminate until the window is closed
* cope with exotic wxwidgets functions to open windows without building=20
a whole app and its own main()...
* be able to communicate with the window in this separate thread
* handle properly thread creating on init, window closing on user's=20
call, window opening again later...

All of these is working nicely, and I really feel that such a terminal=20
can make gnuplot progress.

To resume :

- License issue : wxwidgets is LGPL, so it can be used without=20
restriction inside gnuplot

- Cross-platform : wxwidgets is ported to UNIX through GTK+, through=20
Windows with its native widgets, and even through MacOS (native too)...=20
All these targets would use the same code in gnuplot, and should have=20
identical behaviours !
(For example, there is a Thread class, which allows to use thread=20
without having to take care of the target.)

- All the advantages of a toolkit : it's easy to create toolbars, menus,=20
a statusbar, etc. We can imagine a terminal with its own printing=20
capabilities, (limited but direct) exporting capabilities, a toolbar=20
which makes the mouse mode clearer... I admit that it must not hide=20
gnuplot's own functionnality (printing is well done through the=20
postscript terminal, exporting with corresponding terminals), but I want=20
to try it to see if it can make gnuplot more flexible under special=20
usage (when it's driven by Octave or Maxima for example).

In the next days (weeks ?), I will probably finish implementing gnuplot=20
functionnalities. My base is the X11 terminal for mouse interactivity,=20
and the postscript terminal for rendering. Then it will be time to try=20
other ideas (toolbar ...).

And even if it never get finished or included anywhere, be sure that I=20
learn of things with such a project ! Thanks to the gnuplot authors and=20
to the current community.

Timoth=E9e Lecomte

string variables for font name of set (title|xlabel)

[Shigeharu T. <sh...@ie...>]

shige 06/19 2005
----------------

To use strings variables for the font name of "set title" or "set 
(x|y|z)label", to apply the following patch:

----- From here -----
--- src/set.c.ORG	Sun Jun 19 20:57:17 2005
+++ src/set.c	Sun Jun 19 20:45:15 2005
@@ -3832,9 +3832,22 @@
 
 	if (almost_equals(c_token, "f$ont"))  {
 	    ++c_token;
+#ifdef GP_STRING_VARS
+	    if (isstring(c_token) || isstringvar(c_token)) {	   
+		struct value a;
+		STRING_RESULT_ONLY = TRUE;
+		(void) const_express(&a);
+		STRING_RESULT_ONLY = FALSE;
+		if (a.type == STRING) {
+		    strncpy(label->font, a.v.string_val, MAX_LINE_LEN);
+		    gpfree_string(&a);
+		} else
+		  int_error(c_token,"expecting font");
+#else
 	    if (isstring(c_token)) {
 		quote_str(label->font, c_token, MAX_LINE_LEN);
 		c_token++;
+#endif
 	    } else
 		int_error(c_token,"expecting font");
 	    continue;
----- To here -----

I wish it will be included in CVS tree and I hope the font name 
of "set label" would be allowed to be specified by string 
variables.

+========================================================+
 Shigeharu TAKENO     NIigata Institute of Technology
                       kashiwazaki,Niigata 945-1195 JAPAN
 sh...@ie...   TEL(&FAX): +81-257-22-8161
+========================================================+

Re: Plotting in real time

[Petr M. <mi...@ph...>]

> > Has anyone any other alternative than feeding gnuplot a "plot
> > 'myfile.txt' with lines" every update interval?

plot '-'
1 2
2 4
e

> > I tried feeding data into gnuplot via a fifo but graphing didn't start
> > until the fifo was closed on the  writing end. Is there a way around
> > that?

fflush(out);

---
PM