simultaneous equations in gretl
by Allin Cottrell
Hello all,
Up till now, gretl has supported only three-stage least squares for
estimating systems of simultaneous equations. Now, in CVS and in
the win32 snapshot available from sourceforge, I have added support
for FIML and LIML (along with automatic testing of over-identifying
restrictions). I have also added the Hansen-Sargan test for
over-identifying restrictions, for 3SLS and SUR.
I'm looking for help in testing these new additions. I can
replicate the benchmark results for that workhorse, Klein's "Model
1", but I'm finding it hard to find other benchmarks to test
against. So if anyone has datasets on which they've estimated
equation systems using FIML or LIML with other programs, I'd be very
grateful if you could either:
* See if you can replicate the results using gretl, and let me know
what happens; or
* Send me a copy of the dataset and the output from the other
program, so I can try the replication.
Thanks very much for any assistance.
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
20 years, 6 months
Re: [Gretl-users] Build just gretl library?
by jack
On Mon, 16 May 2005, Skip Montanaro wrote:
> I've been asked to install Gretl on our Solaris environment at work. We're
> missing a number of application-level dependencies. I'm really only
> interested in the gretl library itself, as I hope to wrap it for use in
> Python.
DISCLAIMER: I know practically nothing about Python
Having said this, it would be *extremely* interesting to see examples of
gretl library functions invoked by a Python script.
Skip, would you mind posting something to this effect once you've got it
working?
Riccardo `Jack' Lucchetti
Dipartimento di Economia
Università di Ancona
jack(a)dea.unian.it
http://www.econ.unian.it/lucchetti
20 years, 9 months
gretl on freebsd
by Valerio Daelli
Hello
did anyone manage to install gretl on freebsd?
I am trying and I have some problems.
The biggest is that I can't install LAPACK on freebsd.
Is LAPACK really necessary for gretl?
Tanks a million.
Valerio Daelli
20 years, 9 months
Build just gretl library?
by Skip Montanaro
I've been asked to install Gretl on our Solaris environment at work. We're
missing a number of application-level dependencies. I'm really only
interested in the gretl library itself, as I hope to wrap it for use in
Python. Is it possible to get past the configure step without all the other
dependencies?
Thanks,
--
Skip Montanaro
skip(a)pobox.com
20 years, 9 months
VAR impulse response functions in GUI
by jack
Another one-liner: the following patch makes the GUI frontend to IRFs
conform to a previous "set horizon" command. It shouldn't break anything,
but you never know :-)
Index: gui2/gui_utils.c
===================================================================
RCS file: /cvsroot/gretl/gretl/gui2/gui_utils.c,v
retrieving revision 1.222
diff -u -r1.222 gui_utils.c
--- gui2/gui_utils.c 2 May 2005 20:52:56 -0000 1.222
+++ gui2/gui_utils.c 3 May 2005 10:27:11 -0000
@@ -2898,7 +2898,7 @@
targ = GPOINTER_TO_INT(gtk_object_get_data(GTK_OBJECT(w), "targ"));
#endif
- err = gretl_var_plot_impulse_response(var, targ, shock, 0, datainfo);
+ err = gretl_var_plot_impulse_response(var, targ, shock, get_VAR_horizon(), datainfo);
if (!err) {
register_graph();
Riccardo `Jack' Lucchetti
Dipartimento di Economia
Università di Ancona
jack(a)dea.unian.it
http://www.econ.unian.it/lucchetti
20 years, 10 months
Fractional differencing (yet again)
by jack
This patch fixes a couple of issues with fracdiff: first, the first
observation needn't be discarded, cause fractional differencing is
performed by truncating what is an infinite polynomial anyway. Second,
fracdiff now deals more nicely with missing values at the beginning of a
sample.
Index: lib/src/generate.c
===================================================================
RCS file: /cvsroot/gretl/gretl/lib/src/generate.c,v
retrieving revision 1.234
diff -u -r1.234 generate.c
--- lib/src/generate.c 2 May 2005 14:55:26 -0000 1.234
+++ lib/src/generate.c 3 May 2005 11:15:59 -0000
@@ -3167,8 +3167,6 @@
DPRINTF(("Doing get_fracdiff, with d = %g\n", d));
- if (t1 == 0) t1 = 1;
-
err = series_adjust_t1t2(y, &t1, &t2);
if (err) {
return E_DATA;
@@ -3185,7 +3183,7 @@
}
for (dd=1; dd<=T && fabs(phi)>TOL; dd++) {
- for (t=dd; t<=t2; t++) {
+ for (t=t1+dd; t<=t2; t++) {
if (t >= t1) {
diffvec[t] += phi * y[t - dd];
}
Riccardo `Jack' Lucchetti
Dipartimento di Economia
Università di Ancona
jack(a)dea.unian.it
http://www.econ.unian.it/lucchetti
20 years, 10 months
KPSS in GUI console
by jack
Another one-liner. Trivial patch to enable the kpss command in the gui
console (it was my fault). Allin, could you apply this?
Index: library.c
===================================================================
RCS file: /cvsroot/gretl/gretl/gui2/library.c,v
retrieving revision 1.425
diff -u -r1.425 library.c
--- library.c 30 Apr 2005 17:16:18 -0000 1.425
+++ library.c 2 May 2005 12:34:49 -0000
@@ -5448,6 +5448,7 @@
case GRAPH:
case HURST:
case INFO:
+ case KPSS:
case LABEL:
case LABELS:
case LAGS:
Riccardo `Jack' Lucchetti
Dipartimento di Economia
Università di Ancona
jack(a)dea.unian.it
http://www.econ.unian.it/lucchetti
20 years, 10 months
Slight change in fracdiff
by jack
Allin,
First, thanks for all the goodies recently appeared in CVS.
Please apply the following. The fractional difference operator is
perfectly defined for |d| <= 1; the limiting case d=-1 corresponds to
cumulating a series. Fractionally differencing for d>1 can be accomplished
by "normal" differencing first for the integral part, and then applying
\Delta^d for the fractional part.
Index: lib/src/generate.c
===================================================================
RCS file: /cvsroot/gretl/gretl/lib/src/generate.c,v
retrieving revision 1.232
diff -u -r1.232 generate.c
--- lib/src/generate.c 29 Apr 2005 14:07:05 -0000 1.232
+++ lib/src/generate.c 2 May 2005 10:47:24 -0000
@@ -912,7 +912,7 @@
return NULL;
}
- if (*param < 0.0) { /* ?? */
+ if (fabs(*param) > 1.0) {
genr->err = 1;
return NULL;
}
Riccardo `Jack' Lucchetti
Dipartimento di Economia
Università di Ancona
jack(a)dea.unian.it
http://www.econ.unian.it/lucchetti
20 years, 10 months
