Re: [Gretl-users] wishes list
by a b
Allin Cottrell a écrit :
On Fri, 1 Jun 2007, Hélio Guilherme wrote:
But we really need some more
gretlers who are in a position to make the gretl wish list
_shrink_! (That is, by implementing some of the items on the
list.)
If you know of any graduate students with programming expertise
who might be interested in contributing to gretl, please try to
encourage them to do so!
I have no grasp on C programming but I used to program in Visual Basic a (rudimenatry) econometric software that I went proud of. I gave up when I knew about Gretl! Anyway, could you or anyone else recommand me any good book on C/C++ programming (in english, french or italian) or any other web resources, it should be a pleasure to invest my spare time (and much more than that) to be of any help in contributing to Gretl! It really deserves it!
Thanks once again, for having offered us a useful and great softawre!
Artur
---------------------------------
Stockage illimité de vos mails avec Yahoo! Mail. Changez aujourd'hui de mail !
19 years
System estimation
by Franck Nadaud
Hello listers ! greetings from Paris !
Let me ask a somewhat technical question about system estimation in GRETL.
I am currently working on AIDS models to be estimated as sytems.
So I use the command :
system
...
...
end system
with restrictions inside. However those restrictions imply a singular
covariance matrix. In other pacakges (SAS for example) estimation is done by
deleting one equation, and the missing coefficient is recovered by the
restrictions.
I would like to know how GRETL handles such cases, and more specifically whith
which methods.
After some trials, it seems that GRETL handles quite well, doing neatly all
the computation and printing results as if there was no singularity problem.
Is it right ?
I would like to know if it is possible in system estimation to recover all the
coefficients and standards errors, because i have to compute the AIDS
elasticities.
thanx a lot and forward to read you
cheers
Franck
--
Franck Nadaud
Economiste
CIRED
UMR 8568 CNRS - EHESS
45 bis avenue de la Belle Gabrielle
94736 Nogent-sur-Marne Cedex
TEL 33-1-43-94-73-94
FAX: 33-1-43-94-73-70
MOB: 06-07-39-92-75
France
19 years
structural VARs
by Summers, Peter
Fellow gretl-users,
Do any of you have any experience using gretl for structural VAR work?
On a related note, is there a way to get impulse responses from a
non-choleski decomposition? I'm thinking of something along the lines of
the "impulse" command in RATS: impulse(vcv=blah) ... where 'blah' is my
covariance matrix.
Any thoughts, comments & suggestions would be most welcome!
PS
===============================
Dr. Peter Summers
Assistant Professor
Department of Economics & Geography
Texas Tech University
===============================
19 years
Ang. [Gretl-users] Re: Gretl error [disabling NLS]
by andreas.rosenblad＠ltv.se
cottrell(a)wfu.edu @ INTERNET skrev 2007-08-24 17:17:16 :
> I'll have to think about this. Either I should go through the code
> and fix everything for a non-nls build, or I should remove
> the "--disable-nls" configure option.
Remove it. I do not see any need for it.
Andreas
19 years
System estimation (II)
by Franck Nadaud
Hi again !
Another (quick) question:
in the estimate command, is the switch "sur" for Seemingly Unrelated
Regression functional ? Because when i write it in my script the switch do not
get in bold red so I wonder if it is ok or not ?
a ciao + !
Franck
19 years
gretl CVS update
by Allin Cottrell
I was afraid that yesterday's Windows snapshot might crash on
accessing native gretl databases. Well, it did. But that
should be fixed in the snapshot I have just put up.
Allin Cottrell
19 years
Re: Gretl error [disabling NLS]
by Allin Cottrell
On Fri, 24 Aug 2007, Olvar Bergland wrote:
> I was attempting to compile Gretl version 1.6.5 on a Linux
> (Gentoo) box with NLS turned off. However, I get this error
> message:
>
> ../lib/src/interact.c: In function 'parseopt':
> ../lib/src/interact.c:2450: error: expected expression before 'else'
Thanks, Olvar. I'm cc'ing this reply to the gretl-users list
because it might be of interest there.
I haven't tried building gretl with --disable-nls in a long time,
and I see there are quite a few errors. I fixed the one you
mention, but then many more appeared.
I'll have to think about this. Either I should go through the
code and fix everything for a non-nls build, or I should remove
the "--disable-nls" configure option. Right now, I'm inclined
towards the second option. With current GTK being UTF-8 based,
I'm afraid that even if I got non-nls compilation to work, one
might end up with a broken program.
It used to be that most free software that supported NLS offered a
"disable-nls" switch, but I'm not seeing that so much these days,
and maybe there's a good reason for it!
Any further thoughts on this?
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
19 years
gretl CVS update
by Allin Cottrell
There are a few new things in CVS (and the Windows snapshot) in
response to points made by Ignacio a while back. Here we go:
Databases in GUI:
You can now select multiple series in a database window, using the
usual shift and ctrl modifiers. Then the Display, Graph and
Import menu items will work on all the selected series.
Two things to note: (1) A multiple selection can only include
series with a common frequency. (2) When multiple series are
selected, the mechanism of dragging the series to the main gretl
window doesn't work: this is a long-standing GTK bug. You have to
use /Series/Import from the menu-bar or "Import" from the pop-up
menu.
There is also a new menu item, Delete, which deletes selected
series from the database. This item is available only for native
gretl databases to which the user has write access. [See note
below for MS Windows.]
Databases in CLI:
The "delete" command now has a "--db" option, which allows you to
delete specified series (given by name) from a native gretl
database to which you have write access. You need to use the
"open" command first, as in:
open mydb.bin
delete x15 y32 --db
New "dataset" command:
This subsumes the old commands "addobs" and "transpos", and adds
options to compact or expand the dataset (i.e. lower or increase
the data frequency) which were previously available only via the
GUI menus. Please see the new help item for "dataset", but here's
how it works by example:
dataset addobs 24
dataset compact 1
dataset compact 4 sum
dataset expand 12
dataset tranpose
The command words "addobs" and "transpos" are still available as
aliases, but they're gone from the manual and should be considered
deprecated.
[Note on Windows: It is extaordinarily difficult to find out if
the current user has write access to a given file, involving calls
to 5 win32 API functions with up to 8 arguments per function. I
think I have this right but haven't had time to test. Please be
aware: it's possible the current snapshot will crash when you try
to open a native database -- and gretl tests to see if you have
write access to it.]
Allin.
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
19 years
testing restrictions on VECMs (2)
by Allin Cottrell
I have now fixed various breakages in CVS gretl, caused by recent
changes. I've also updated the documentation for testing
restrictions on VECMs, and have added three example scripts (in
respect of examples from Johansen, Boswijk and Doornik).
See http://www.ecn.wfu.edu/~cottrell/gretl/
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
19 years
