10:15 a.m.
I was teaching someone how to use gretl for research and while showing
the various plot options, the first thing she noticed and asked about
was whether it is possible to have "bar graphs". I think it would
really be useful if we could select "bars" among the various line
types such as points, impulses and steps. This would really open up
for some very nice looking plots involving combinations of lines and
vertical bars especially useful for time-series analysis.
Cheers
Talha
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
6:55 p.m.
On Thu, 27 Nov 2008, Talha Yalta wrote:
I was teaching someone how to use gretl for research and while
showing
the various plot options, the first thing she noticed and asked about
was whether it is possible to have "bar graphs". I think it would
really be useful if we could select "bars" among the various line
types such as points, impulses and steps...
You can use "impulses" (and choose a greater line width if you
wish).
--
Allin Cottrell
Department of Economics
Wake Forest University
8:39 p.m.
> I was teaching someone how to use gretl for research and while
showing
> the various plot options, the first thing she noticed and asked about
> was whether it is possible to have "bar graphs". I think it would
> really be useful if we could select "bars" among the various line
> types such as points, impulses and steps...
You can use "impulses" (and choose a greater line width if you
wish).
Hey this is pretty nice but doesn't quite achieve the effect since the
maximum line width is 6. Plus I don't think most people will figure
out this trick themselves. It would be great if this was an option
where the base width is automatically determined by the horizontal
scale. Would be even better with a black line framing each box (the
line width would be the width of the frame in this case, or, it can be
the relative width of the bar in percentages so that we can have more
than one set of bars).
Thanks to all your efforts, gretl is getting more and more refined
each day. IMHO this is one of those small things that can make a
difference especially in the eyes of newcomers.
Sincerely
Talha
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
11:17 p.m.
On Thu, 27 Nov 2008, Allin Cottrell wrote:
On Thu, 27 Nov 2008, Talha Yalta wrote:
> I was teaching someone how to use gretl for research and while showing
> the various plot options, the first thing she noticed and asked about
> was whether it is possible to have "bar graphs". I think it would
> really be useful if we could select "bars" among the various line
> types such as points, impulses and steps...
You can use "impulses" (and choose a greater line width if you
wish).
Uhmm... maybe I fail to see something really obvious here, but...
Talha: try applying the following patch; would you be happy with the
result?
Allin: are there any undesirable side-effects with this patch I haven't
thought of?
Index: gui2/gpt_dialog.c
===================================================================
RCS file: /cvsroot/gretl/gretl/gui2/gpt_dialog.c,v
retrieving revision 1.109
diff -u -r1.109 gpt_dialog.c
--- gui2/gpt_dialog.c 21 Nov 2008 21:56:51 -0000 1.109
+++ gui2/gpt_dialog.c 27 Nov 2008 23:11:11 -0000
@@ -1926,6 +1926,7 @@
stylist = g_list_append(stylist, "impulses");
stylist = g_list_append(stylist, "dots");
stylist = g_list_append(stylist, "steps");
+ stylist = g_list_append(stylist, "boxes");
vbox = gp_dialog_vbox();
gtk_widget_show(vbox);
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
7:20 a.m.
Talha: try applying the following patch; would you be happy with the
result?
Allin: are there any undesirable side-effects with this patch I haven't
thought of?
Thanks very much but how do I apply it? I have never done this before.
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
9:46 a.m.
On Fri, 28 Nov 2008, Talha Yalta wrote:
> Talha: try applying the following patch; would you be happy with
the result?
>
> Allin: are there any undesirable side-effects with this patch I haven't
> thought of?
Thanks very much but how do I apply it? I have never done this before.
Crash course in diff unified format:
Index: gui2/gpt_dialog.c
===================================================================
RCS file: /cvsroot/gretl/gretl/gui2/gpt_dialog.c,v
retrieving revision 1.109
diff -u -r1.109 gpt_dialog.c
--- gui2/gpt_dialog.c 21 Nov 2008 21:56:51 -0000 1.109
+++ gui2/gpt_dialog.c 27 Nov 2008 23:11:11 -0000
@@ -1926,6 +1926,7 @@
stylist = g_list_append(stylist, "impulses");
stylist = g_list_append(stylist, "dots");
stylist = g_list_append(stylist, "steps");
+ stylist = g_list_append(stylist, "boxes");
vbox = gp_dialog_vbox();
gtk_widget_show(vbox);
The above is called a patch. It tells you that the file that needs
modifications in the gretl source tree is "gui2/gpt_dialog.c" (it says so
on top). In this case, the only modification is to add the line
stylist = g_list_append(stylist, "boxes");
"Adding" is where the "+" in the patch comes from. If there had been
any
lines to delete from the source, you'd have had a minus sign.
You can find the exact point in the source where to do add the line by
looking for "@@". In this case, the patch shows you the point where the
adding goes by surrounding it with a few context lines. The first line
you're shown is line 1926.
So, in practice you may apply the patch by
- making sure your source tree is up to date
- opening the file gui2/gpt_dialog.c with any editor you like
- navigating to line 1926
- adding the appropriate line into the appropriate place
- building gretl as per usual
On a unix system, most of these steps can be automated via a utility
called "patch". See "man patch" for details.
HTH,
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
12:28 p.m.
Crash course in diff unified format:
Thanks very much Jack. I am away from my computer now but will try it
ASAP, hopefully tomorrow. To turn this into a patch file, all I do is
to copy all the lines into a text file right? Do I need a specific
file name or extension?
Cheers
Talha
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
12:32 p.m.
On Fri, 28 Nov 2008, Talha Yalta wrote:
> Crash course in diff unified format:
>
Thanks very much Jack. I am away from my computer now but will try it
ASAP, hopefully tomorrow. To turn this into a patch file, all I do is
to copy all the lines into a text file right?
yes
Do I need a specific file name or extension?
no, but the .diff or .patch extensions are customary.
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
12:04 p.m.
New subject: Feature request: Nelder-Mead optimization algorithm
Hi,
It would be nice if the Nelder-Mead optimization algorithm could be
included in gretl, in addition to the already available BFGS algorithm.
R has C code for Nelder-Mead that probably could be included in gretl
without to much work (e.g. an old version here
http://www.google.com/codesearch/p?hl=en#p2xSPglSHSE/R-2.0.1
/src/main/optim.c&q=Nelder-Mead%20lang:c ).
Best regards
Andreas
9:21 p.m.
I tried it and it works nicely. I don't know how you did all this by
adding just one line. That said, well, ehem, Jack could you please
please do it so that the boxes are filled? This way we can have
line-box combination graphs looking much nicer. The frame could be
black for all boxes or, preferably, one level darker. Essentially, I
am talking about the auto-generated plot in gretl's normality test of
the residual. It would be very neat if we could have this very same
effect in the other plots.
And Allin, the items lines,points,linespoints,impulses,... are not
marked for translation. Could you add these to the pot file please? It
doesn't look nice when things are part translated and part in English.
Sincerely,
Talha
On Sat, Nov 29, 2008 at 6:58 PM, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Fri, 28 Nov 2008, Riccardo (Jack) Lucchetti wrote:
[patch adding "boxes" style]
> Allin: are there any undesirable side-effects with this patch I haven't
> thought of?
Hmm, no, it seems to be OK.
Allin.
_______________________________________________
Gretl-devel mailing list
Gretl-devel(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-devel
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
10:09 p.m.
On Sat, 29 Nov 2008, Talha Yalta wrote:
And Allin, the items lines,points,linespoints,impulses,... are
not marked for translation. Could you add these to the pot file
please? It doesn't look nice when things are part translated and
part in English.
I'm not sure that's a good idea: these elements are passed
verbatim to gnuplot. We could "untranslate" them on the way to
gnuplot, but using the proper gnuplot names for these styles might
make it easier for people to learn how to use gnuplot for
themselves.
Allin.
8:40 p.m.
I disagree because I guess like most people it didn't even occur to me
that those options were infact gnuplot commands. I don't know what
other parameters are involed for those plots but if I need a more
hands on approach, I will have to read the gnuplot documentation
anyway. Meanwhile, for the majority of users those parameters look
like a bunch of untranslated strings, possibly due to a lousy
translation job. If other people do not disagree, and if it will not
be too difficult to implement, could we go with the untranslation
layer for this one please?
Sincerely
Talha
On Sun, Nov 30, 2008 at 12:09 AM, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Sat, 29 Nov 2008, Talha Yalta wrote:
> And Allin, the items lines,points,linespoints,impulses,... are
> not marked for translation. Could you add these to the pot file
> please? It doesn't look nice when things are part translated and
> part in English.
I'm not sure that's a good idea: these elements are passed
verbatim to gnuplot. We could "untranslate" them on the way to
gnuplot, but using the proper gnuplot names for these styles might
make it easier for people to learn how to use gnuplot for
themselves.
Allin.
_______________________________________________
Gretl-devel mailing list
Gretl-devel(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-devel
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
5814
days inactive
5868
days old
15 comments
5 participants
participants (5)
-
Alexander B. Gedranovich -
Allin Cottrell -
andreas.rosenblad＠ltv.se -
Riccardo (Jack) Lucchetti -
Talha Yalta