5:58 p.m.
Hello all,
We're planning to release gretl 1.9.6 before long -- probably a
couple of weeks from now -- and in the meantime we'd much appreciate
it if you could help us hunt down bugs, so that we can make the best
possible release.
If you have old gretl scripts to hand, please try running them and
check that they still work. And if you have a record of results,
please check that the current results agree with the old ones.
Also -- since checking the gretl GUI for breakage is not so easy as
checking scripts -- please see if you can find time to do things
using the graphical interface and look out for anything wrong.
Send your reports of bugs to this list, or use the bug-tracker at
Sourceforge if you prefer:
http://sourceforge.net/tracker/?group_id=36234&atid=416803
For now, however, please hold back on feature requests -- we're
focused on finding bugs.
Please bear in mind a couple of important points:
1) Your reports will be much more valuable if you're using gretl
that you have build yourself from current CVS, or else the current
snapshot for MS Windows
http://gretl.sourceforge.net/win32/
or Mac OS X.
http://gretl.sourceforge.net/osx.html
Many bugs in the last release (gretl 1.9.5) have already been fixed.
See http://gretl.sourceforge.net/ChangeLog.html .
2) Some older gretl scripts will not work because the scripts
thenselves contain bugs. Current gretl catches (for example) the
syntax error of writing
scalar x = 2*(3 + 4)/y)
which has an extraneous trailing ')'. This was fixed just today
(2011-09-04). I was surprised at how many scripts in my collection
exemplified this error.
Thanks!
Allin Cottrell
6:02 a.m.
New subject: bug-hunting: underlined letters
(gretl in english downloaded from current CVS and running in
Ubuntu-linux 10.04)
With respect to the GUI, I have seen from some time that the underlined
letters in the menus seems not to work correctly.
They only work for the main menu, but not for subelements of the main
menu. For example, if I press ALT+F the "File" entry is opened, but then
pressing on ALT+N the "New data set" entry is not opened (the same
occurs with ALT+ any other).
(As I write this I see that as Spanish translator I have been totally
inconsistent with the underlined letters in Spanish. There are several
repetitions of letters even in the main menu. I will do a review of all
of them and will correct that soon.)
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.ea3.ehu.es
9 a.m.
New subject: bug-hunting: underlined letters
On Mon, 5 Sep 2011, Ignacio Diaz-Emparanza wrote:
(gretl in english downloaded from current CVS and running in
Ubuntu-linux 10.04)
With respect to the GUI, I have seen from some time that the underlined
letters in the menus seems not to work correctly.
They only work for the main menu, but not for subelements of the main
menu. For example, if I press ALT+F the "File" entry is opened, but then
pressing on ALT+N the "New data set" entry is not opened (the same
occurs with ALT+ any other).
Once you're in the File menu via Alt-F, just press 'N' for new
dataset (not Alt-N), and similarly for other menus and
entries. This is just how GTK works.
Allin
9:17 a.m.
New subject: bug-hunting: underlined letters
El 05/09/11 15:00, Allin Cottrell escribió:
On Mon, 5 Sep 2011, Ignacio Diaz-Emparanza wrote:
> (gretl in english downloaded from current CVS and running in
> Ubuntu-linux 10.04)
>
> With respect to the GUI, I have seen from some time that the underlined
> letters in the menus seems not to work correctly.
> They only work for the main menu, but not for subelements of the main
> menu. For example, if I press ALT+F the "File" entry is opened, but then
> pressing on ALT+N the "New data set" entry is not opened (the same
> occurs with ALT+ any other).
Once you're in the File menu via Alt-F, just press 'N' for new
dataset (not Alt-N), and similarly for other menus and
entries. This is just how GTK works.
Allin
Umm, ... this behaviour is different as in other programs. For example,
in Thunderbird or Openoffice in the second stage
pressing ALT+N (the letter of the submenu) works as well.
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.ea3.ehu.es
10:02 a.m.
New subject: bug-hunting: underlined letters
On Mon, 5 Sep 2011, Ignacio Diaz-Emparanza wrote:
El 05/09/11 15:00, Allin Cottrell escribió:
> On Mon, 5 Sep 2011, Ignacio Diaz-Emparanza wrote:
>
>> (gretl in english downloaded from current CVS and running in
>> Ubuntu-linux 10.04)
>>
>> With respect to the GUI, I have seen from some time that the underlined
>> letters in the menus seems not to work correctly.
>> They only work for the main menu, but not for subelements of the main
>> menu. For example, if I press ALT+F the "File" entry is opened, but
then
>> pressing on ALT+N the "New data set" entry is not opened (the same
>> occurs with ALT+ any other).
> Once you're in the File menu via Alt-F, just press 'N' for new
> dataset (not Alt-N), and similarly for other menus and
> entries. This is just how GTK works.
Umm, ... this behaviour is different as in other programs. For example,
in Thunderbird or Openoffice in the second stage
pressing ALT+N (the letter of the submenu) works as well.
I see that you're right. I don't know why that should be. I'll
look into it.
Allin
10:22 a.m.
New subject: Choice of Cointegration and adjustment vectors in a multivariate VAR
Dear Grelt-users,
I'll appreciate if someone could give me a hint on the
choice of cointegration and adjustment vectors in a multivariate VAR. To be more specific,
results of my Johansen cointegration test reflects a cointegration rank of above
2(r>2). This implies, running a corresponding VECM model with thesame deterministic
components as in the Johansen test should provide more than two cointegration vectors and
Error correction terms(adjustment vectors). I don't know what criterion to use in
selecting the right cointegration equation and speed of adjustment; as hitherto, I've
been used to applying only bivariate VARs.
Thank you
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
10:28 a.m.
New subject: bug-hunting: underlined letters
On 05-09-2011, at 16:02, Allin Cottrell wrote:
On Mon, 5 Sep 2011, Ignacio Diaz-Emparanza wrote:
> El 05/09/11 15:00, Allin Cottrell escribió:
>> On Mon, 5 Sep 2011, Ignacio Diaz-Emparanza wrote:
>>
>>> (gretl in english downloaded from current CVS and running in
>>> Ubuntu-linux 10.04)
>>>
>>> With respect to the GUI, I have seen from some time that the underlined
>>> letters in the menus seems not to work correctly.
>>> They only work for the main menu, but not for subelements of the main
>>> menu. For example, if I press ALT+F the "File" entry is opened, but
then
>>> pressing on ALT+N the "New data set" entry is not opened (the same
>>> occurs with ALT+ any other).
>> Once you're in the File menu via Alt-F, just press 'N' for new
>> dataset (not Alt-N), and similarly for other menus and
>> entries. This is just how GTK works.
>
> Umm, ... this behaviour is different as in other programs. For example,
> in Thunderbird or Openoffice in the second stage
> pressing ALT+N (the letter of the submenu) works as well.
I see that you're right. I don't know why that should be. I'll look into it.
I've been experimenting in Ubuntu 10.04 in VirtualBox.
The behaviour of programs seems to vary.
In Firefox you use the Alt+letter to go to a main menu item and then
- you can keep the Alt key depressed and press an underlined letter
- or you can let go of the Alt key and and press an underlined letter.
The Gnome Commander on the other hand wants you to let go of the Alt key when you go for a
submenu item. Nautilus behaves in the same way.
And Gretl of course.
Identical behaviour in Xubuntu 11.04.
Berend
2:56 p.m.
New subject: bug-hunting: underlined letters
On Mon, 5 Sep 2011, Berend Hasselman wrote:
I've been experimenting in Ubuntu 10.04 in VirtualBox.
The behaviour of programs seems to vary.
In Firefox you use the Alt+letter to go to a main menu item and then
- you can keep the Alt key depressed and press an underlined letter
- or you can let go of the Alt key and and press an underlined letter.
The Gnome Commander on the other hand wants you to let go of
the Alt key when you go for a submenu item. Nautilus behaves
in the same way. And Gretl of course.
I'm seeing the same here on native linux.
As an experiment I wrote a minimal test program which just
opens a plain GTK window with a menu. There are at least three
different APIs for constructing a menu in GTK (though maybe
they all come to the same thing internally). Anyway, I tried
building the menu in three ways and in all cases it behaved
like the ones in gretl and gnome apps, as Berend describes.
My tentative conclusion is that apps such as Firefox and
LibreOffice, where both Alt-key and plain key can be used to
select a submenu item with an underlined letter, must be doing
something special to get that effect. In which case I don't
think it's worth bothering with.
Allin
3:45 p.m.
New subject: bug-hunting: underlined letters
On 05-09-2011, at 20:56, Allin Cottrell wrote:
On Mon, 5 Sep 2011, Berend Hasselman wrote:
> I've been experimenting in Ubuntu 10.04 in VirtualBox.
> The behaviour of programs seems to vary.
>
> In Firefox you use the Alt+letter to go to a main menu item and then
> - you can keep the Alt key depressed and press an underlined letter
> - or you can let go of the Alt key and and press an underlined letter.
>
> The Gnome Commander on the other hand wants you to let go of
> the Alt key when you go for a submenu item. Nautilus behaves
> in the same way. And Gretl of course.
I'm seeing the same here on native linux.
As an experiment I wrote a minimal test program which just
opens a plain GTK window with a menu. There are at least three
different APIs for constructing a menu in GTK (though maybe
they all come to the same thing internally). Anyway, I tried
building the menu in three ways and in all cases it behaved
like the ones in gretl and gnome apps, as Berend describes.
My tentative conclusion is that apps such as Firefox and
LibreOffice, where both Alt-key and plain key can be used to
select a submenu item with an underlined letter, must be doing
something special to get that effect. In which case I don't
think it's worth bothering with.
I agree.
I've just checked to see what happens on Windows.
It's also a mixed bag.
In TextPad (editor) one must keep Alt depressed when accessing sub menu items.
Total Commander behaves like Firefox on Ubuntu, so you don't have to keep depressing
the Alt key when accessing a sub menu item with an underlined letter.
Firefox on Windows behaves as Firefox on Ubuntu.
Berend
5:50 a.m.
New subject: bug-hunting: underlined letters
Il giorno lun, 05/09/2011 alle 14.56 -0400, Allin Cottrell ha scritto:
On Mon, 5 Sep 2011, Berend Hasselman wrote:
> I've been experimenting in Ubuntu 10.04 in VirtualBox.
> The behaviour of programs seems to vary.
>
> In Firefox you use the Alt+letter to go to a main menu item and then
> - you can keep the Alt key depressed and press an underlined letter
> - or you can let go of the Alt key and and press an underlined letter.
>
> The Gnome Commander on the other hand wants you to let go of
> the Alt key when you go for a submenu item. Nautilus behaves
> in the same way. And Gretl of course.
I'm seeing the same here on native linux.
As an experiment I wrote a minimal test program which just
opens a plain GTK window with a menu. There are at least three
different APIs for constructing a menu in GTK (though maybe
they all come to the same thing internally). Anyway, I tried
building the menu in three ways and in all cases it behaved
like the ones in gretl and gnome apps, as Berend describes.
My tentative conclusion is that apps such as Firefox and
LibreOffice, where both Alt-key and plain key can be used to
select a submenu item with an underlined letter, must be doing
something special to get that effect. In which case I don't
think it's worth bothering with.
Yes, exactly. In fact, of all the Linux apps mentioned, Firefox and
LibreOffice are the only two not natively nor directly gtk-based: the
first uses XUL, the second has its own widgets toolkit.
Pietro
10:36 a.m.
New subject: bug-hunting: underlined letters
El 05/09/11 20:56, Allin Cottrell escribió:
...
As an experiment I wrote a minimal test program which just
opens a plain GTK window with a menu. There are at least three
different APIs for constructing a menu in GTK (though maybe
they all come to the same thing internally). Anyway, I tried
building the menu in three ways and in all cases it behaved
like the ones in gretl and gnome apps, as Berend describes.
OK. So this is not a bug, I did not know how to use correctly the
underlined letters. I have learned now, thanks.
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.ea3.ehu.es
10:57 a.m.
Hello Allin,
This is not actually a bug but may be you'll find it useful to point out.
I'm currenty handling a large census database and:
1. I observed some limitations through the GUI:
a. adding new observations is limited to 100 00 obs;
b. creating a new data set is limited to 1 000 000 obs;
Otherwise, there's no limits to the "addobs" and
"nulldata"
commands in the CLI
2. Plotting the frequency distribution of a constant series X (GUI) pops
up a very clear error message " 'X' is a constant". Instead, the
"freq"
command (CLI) ends up with 3 different error messages not so easy to
figure out.
In the same run, the boxplot command has the same behavior (GUI-CLI)
although the error message should perhaps, be more intuitive (like
frequency's "'X' is a constant" for example instead of
"C:\Program
Files\gretl\wgnuplot.exe"
"C:\Users\Artur\AppData\Roaming\gretl\gpttmp.i03604": exit code 1").
3. While executing commands in the CLI - specially loop iterations -
gretl seems to shut down but actually it doesn't.
Would be possible to show something like a progress bar (in fact, we
used to have a stuff like that in the script window) just to tell the
user that something is still going on?
Cheers,
artur
4:53 p.m.
On Tue, 6 Sep 2011, artur bala wrote:
This is not actually a bug but may be you'll find it useful to
point out.
Thanks, Artur. Yes, your comments are useful.
I'm currenty handling a large census database and:
1. I observed some limitations through the GUI:
a. adding new observations is limited to 100 00 obs;
b. creating a new data set is limited to 1 000 000 obs;
Otherwise, there's no limits to the "addobs" and
"nulldata"
commands in the CLI
I'm just trying to guard against the GUI becoming "frozen" due
to an excessive memory request. I guess I could raise the
limits but put up an "Are you sure?" dialog for values greater
than the current limits
2. Plotting the frequency distribution of a constant series X (GUI)
pops
up a very clear error message " 'X' is a constant". Instead, the
"freq"
command (CLI) ends up with 3 different error messages not so easy to
figure out.
You're right, that was a mess. It should be cleaned up now in
CVS.
In the same run, the boxplot command has the same behavior (GUI-CLI)
although the error message should perhaps, be more intuitive (like
frequency's "'X' is a constant" for example instead of
"C:\Program
Files\gretl\wgnuplot.exe"
"C:\Users\Artur\AppData\Roaming\gretl\gpttmp.i03604": exit code 1").
Same answer as above.
3. While executing commands in the CLI - specially loop iterations -
gretl seems to shut down but actually it doesn't.
Would be possible to show something like a progress bar (in fact, we
used to have a stuff like that in the script window) just to tell the
user that something is still going on?
We have something like that in the script window, namely a
"spinner" which should start going round when a time-consuming
script is being executed. It's not always possible for this to
move in real time because it's updated just whenever gretl
"comes up for air", but in my experiments it works reasonably
well.
Allin
11:15 p.m.
New subject: Interpreting Impulse-response graphs
Hi all,
Could someone gist me on how to interprete impulse-response graphs. This is with
respect to Error Correction Terms(ECTs) in a VECM.
Thanks in advance
2:11 a.m.
New subject: Interpreting Impulse-response graphs
On Wed, 7 Sep 2011, Anutechia Asongu wrote:
Hi all, Could someone gist me on how to interprete
impulse-response graphs. This is with respect to Error Correction
Terms(ECTs) in a VECM.
http://books.google.com/books?id=muorJ6FHIiEC&pg=PA262&source=gbs...
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
5:31 a.m.
New subject: bug-hunting time - scalars window
dear allin,
there's a strange behaviour of the scalars window (see a screen capture
/http://www.artbala.net/scalars.jpg/)
I run a script which opens a dataset and produces some scalars. Then I
open the scalars windows to view the output.
Without closing the scalars windows I re-run the script twice by
changing only the dataset filename to be opened and analyzed.
The scalars window is not refreshed and the new scalars are piled up to
the old ones keeping the same names and, some of them showing strange
values. Some new and not requested scalars appear for the first time
(the 'i's in the picture) while some others supposed to be deleted still
appear (the 'essai's in the picture).
best,
artur
11:22 a.m.
New subject: bug-hunting time - scalars window
On Tue, 13 Sep 2011, artur bala wrote:
there's a strange behaviour of the scalars window (see a screen
capture
/http://www.artbala.net/scalars.jpg/)
Thanks, Artur. The problem is that the scalars window is just
not set up to update properly when you run a script: closing
it and reopening it should show correct values.
Obviously this ought to be fixed -- either so it updates
correctly, or it's automatically closed when you run a script.
I'll be offline for a few days, but will take a look at it
when I can.
Allin Cottrell
8:39 a.m.
New subject: gretl_x11 -r
I know that if you call the gretlcli with
gretlcli -r filename.inp
then gretl opens and run this file and then passes control to the user,
but I see that if I try to do the same thing with the GUI client:
gretl_x11 -r filename.inp
the file is opened in the script editor, but not executed (you have to
click over the execute icon). Is there any way that this script may be
automatically executed (without clicking over the icon)?
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.ea3.ehu.es
5:59 p.m.
New subject: bug-hunting time : vecm
Hi,
After running the "hamilton.inp" script I tried to plot the response of
p to s from the the hamilton model menu but
gretl comes out with a "Data error" message . Otherwise, when I checked
the "include bootstrap confidence interval"
everything looks fine.
Best,
artur
8:49 p.m.
New subject: bug-hunting time : vecm
On Tue, 4 Oct 2011, artur bala wrote:
After running the "hamilton.inp" script I tried to plot the
response of
p to s from the the hamilton model menu but
gretl comes out with a "Data error" message . Otherwise, when I checked
the "include bootstrap confidence interval"
everything looks fine.
Thanks, Artur. I can replicate the problem, which (hopefully) means
that I can fix it fairly quickly.
Allin Cottrell
9:04 p.m.
New subject: bug-hunting time : vecm
On Tue, 4 Oct 2011, Allin Cottrell wrote:
On Tue, 4 Oct 2011, artur bala wrote:
> After running the "hamilton.inp" script I tried to plot the response of
> p to s from the the hamilton model menu but
> gretl comes out with a "Data error" message . Otherwise, when I checked
> the "include bootstrap confidence interval"
> everything looks fine.
Thanks, Artur. I can replicate the problem, which (hopefully) means that I
can fix it fairly quickly.
OK, it's fixed in CVS, will be fixed in the snapshots tomorrow.
Allin Cottrell
4751
days inactive
4782
days old
21 comments
7 participants
participants (7)
-
Allin Cottrell -
Anutechia Asongu -
artur bala -
Berend Hasselman -
Ignacio Diaz-Emparanza -
Pietro Battiston -
Riccardo (Jack) Lucchetti