On Mon, 17 Jan 2011, [ISO-8859-1] Alejandro Mosi�o wrote:
> Maybe i was not too much specific last time:
> I have a variable "y" that follows a GARCH(1,1) process. Then, i Gretl i
> garch 1 1 ; y const
> Then i got the result and forecasting the out-of-sample values of y can
> be done in the usual way. However, i'm interested in forecasting the
> out-of-sample variance. I don't know if such a function exists in Gretl.
There is no built-in function to do this, but you can compute a
one-step ahead forecast of the variance from the model data, as
hown in the following example script.
garch 1 1 ; Y
series e = $uhat
series h = $h
dataset addobs 10
a0 = $coeff
a1 = $coeff
b1 = $coeff
series hfc = h
# set future errors to their expectation
e = misszero(e)
# forecast the variance
hfc = a0 + a1 * e(-1)^2 + b1 * hfc(-1)
smpl 1970 ;
print e h hfc --byobs
I know there are functions to compute the mean, median, minimum, maximum and standard deviation for a series. I also know that using the summary command, skewness and excess-kurtosis are also computed. Is there an easy way to capture these last two statistics (I need them for further processing)? Of course, I could write a short function to do this, but as they are already available in Gretl in the summary command…
UNIVERSITEITANTWERPEN | Faculty of Applied Economics (TEW) | Dept. Accounting & Finance
Room S.B.335 | Prinsstraat 13 | B-2000 Antwerp | Belgium
Phone +32 32654163 |Fax +32 32654064
I just ran across a problem trying to modify a data set. If I delete a variable, then do ctrl-s (on windows 7), I get an error saying "s is not the name of a variable." The same thing happens if I select "save data" or "save data as..." from the file menu. In the latter case, the error happens at the end, after going through all the dialog boxes.
My gretl version was a few weeks old, so I downloaded the latest cvs version & it's still there.
A couple of new things for testing:
1) On OS X I've enabled the saving of stuff that is downloaded
from the gretl server (functions, databases, data-file
packages) to ~/Library/Application Support/gretl, as
Berend H suggested. Only minimally checked so far.
2) In the packages for OS X and MS Windows I've updated
gtksourceview (which supports our syntax highlighting) to
version 2.10.5, which means that we can print scripts with
highlighting, if wanted. This seems to work fine on Linux and
OS X. On Windows 7 I'm getting weird results so far (some sort
of units mismatch between GTK and Windows, such that the
"page" appears to be about 1 by 2 inches). I'd be interested
to hear if it works for others.
3) As mentioned in another thread, you can now control the
parameters of a frequency distribution ("freq" command)
via the command line, as you can in the GUI.
How to set other number of class for frequency distribution table?
Prof. Carlos A. S. de Andrade
LAPEA - Laboratório de Pesquisa em Economia Aplicada e Engenharia de
Universidade Federal de Campina Grande.
Centro de Humanidades
Unidade Acadêmica de Economia
I'm estimating an ARMA(2,2) model from generated data using gretl's
exact likelihood. When I forecast 10 periods out from the end of the
sample, the first 2 forecast standard errors are as I would expect.
However, after this the standard error of the prediction is larger
than the standard error of my series. I was under the impression that
this cannot be. Surely, I'm misunderstanding something? The standard
error does approach some limit, though I'm not sure what. Does someone
have a reference for this? I didn't see anything in the gretl user
The series is here, if someone wants to replicate.
I am trying to learn libgretl and have one question: in gretl_VAR(), the
description of the 4th parameter (opt) says "if includes OPT_R, ..", suggesting
that one can combine several flags. Also the source code seems to expect this,
having checks of type "opt & flag".
However, the parameter is of type gretlopt, which is an enum - so that I have to
use exactly one number from the list (at least in C++, not sure about C).
Shouldn't the parameter be int, so that one could add several values from
gretlopt - or did I misunderstood something, as usual?
is it possible to include a model type object into the bundle?
RES <- ols lg const lg(-1) --quiet
foo["model"] = "RES"
Is there a quick way to code the following without individually specifying
each right hand side lag?
Y = Y(-1) Y(-2) X(-1) X(-2)
X = Y(-1) Y(-2) X(-1) X(-2)
I ask this because I have a large model with four lags. So I was wondering
if there is a quick way to do something like the following
Y = Y(-1 to -2) ....
X = Y(-1 to -2) ....
I've been using seemingly unrelated regressions (sur) method and system
Dear Prof. Cottrell, Dear GRETL Users,
Is there a downloadable GRETL subroutine for Maximum Entropy or Minimum Cross-Entropy regression (and if
possible SUR by entropy). If so, would you be so kind to indicate the address?
Thanking you in advance, I send you my best regards from Ankara.