For a simulation study, I need to compute OLS with plain std errors, and
with all versions of heteroskedasticity-robust std errors.
Do I go something like the following? (Essentially I want to know if the
next "set" command on the same thing overrides the previous or whether
the previous needs clearing first)
loop i = 1..100
ols Y Xlist
set hc_version 0
ols Y Xlist -- robust
set hc_version 1
ols Y Xlist -- robust
set hc_version 3a
ols Y Xlist -- robust
Alecos Papadopoulos PhD
Athens University of Economics and Business
Hi gretl users,
there is version 2.0 of the "Threshold_Panel" out now. The package
implements Bruce E. Hansen's panel-threshold model (JoE, 1999).
Apart from some internal refactoring (increasing computational speed
too!) and bug fixes, the user-interface has changed too for improving
the "user-experience". The most important changes are the first two
bullet points in the following changelog. All details can be found in
the help file as always:
2.0 (March 2021)
- Changed API: Rename function SETUP_THRESH() to estimateTpm() and change
- Changed API: Rename function lrplot() by plotTpm() and change
- Replace "stop" by errorif() function.
- Fix bug in get_interaction_terms() function for constructing
- Make use of hdprod() function for computing interaction terms;
results in about +20% gain of speed.
- Internal refactoring and re-formatting (about 5% gain of speed).
- Update help file.
- Improve print-out of tests (formatting)
- Minimum required version is 2020b now.
there's a package with a new name on the package download server:
But it's basically just the old package coint2rec with a new name,
following the name change of the gretl command from "coint2" to
"johansen". (From the help text: "This package provides stability tests
for the cointegrated vector autoregressive model based on recursive
So if you have any scripts that used coint2rec, you just need to change
your "include" line and the function calls accordingly. The earlier top
level function name was driver_coint2recur and received some criticism
for not being very intuitve, so that is now "johansenrec" as well.
But usage is unchanged!
The old package coint2rec will be retired from the server soon, but of
course if you have downloaded it to your local machines it will just
continue to work there.
I would like to save a plot/graph as a .png file using the c/c++ api. In order to do this do I need to use header files from the gui folder. In particular do I need to use gpt_control.h, and session.h? If there are an examples on how to save plot/graph as a .png file that would be great, if not an overview on how to go about doing it would be much appreciated.
When I had my first problems with Gretl under Zoom I asked a rather
knowledgeable IT collegue from my department (the source of the
suggestion to upgrade Gretl in Windows priority settings). He wasn't
very surprised. Apparently Zoom, although nice to use, is rather
rudimentary. For instance, the compression algorithms are rather naive
if compared to MS Teams (which is built upon Skype's engine). So the
bottom line is probably avoiding using it if at all possible (I cannot,
it's the platform of choice of my university - damn it).
best of luck
Professore Ordinario di Statistica Economica
Dip. di Scienze Statistiche
Università di Roma "La Sapienza"
P.le A. Moro 5 - 00185 Roma - Italia
contenute in questo messaggio di posta elettronica sono strettamente
riservate e indirizzate esclusivamente al destinatario. Si prega di non
leggere, fare copia, inoltrare a terzi o conservare tale messaggio se non
si è il legittimo destinatario dello stesso. Qualora tale messaggio sia
stato ricevuto per errore, si prega di restituirlo al mittente e di
cancellarlo permanentemente dal proprio computer.
The information contained
in this e mail message is strictly confidential and intended for the use of
the addressee only. If you are not the intended recipient, please do not
read, copy, forward or store it on your computer. If you have received the
message in error, please forward it back to the sender and delete it
permanently from your computer system.
motivated by some examples shown here earlier by Reynaldo we have
updated the PMG contributed function package to version number 1.0. (It
does pooled mean group estimation for large-T panel datasets.) The main
computational news is that the default initial values used for the
estimation algorithm should be better now for most cases. (However,
coauthor Jack also found an example where the old initialization is
actually better, so as the user you get an --expert-- choice and can try
both if needed.)
For scripting there's also a slightly simpler interface function which
is just called PMG(). See the included pdf help file of the package.
As always, you can download and install the new package version directly
from within gretl.
I am not sure whether this is a bug or following gretl's specs: Putting
a string-valued series into a bundle and creating a new series 'foo'
using this 'bundled' series yields a numeric series 'foo'. I would have
expected the original string-valued series.
See the following example. The output is from latest git version on
set verbose off
open grunfeld.gdt -q
bundle Names = _(names = firmname) # put series 'firmname' into a
series foo = Names.names # construct new series from bundle
print firmname foo -o --range=1:2
1:01 General Motors 1
1:02 General Motors 1