for a given panel dataset with embedded information about the time
dimension structure I want to obtain the periodicity in the time
direction; for example the value 4 if the data are quarterly (for
several cross units, of course, because we're in a panel). I know that
the $pd accessor doesn't work here because in a panel context it gives
me the total length of the time dimension (as documented).
I'm wondering whether there is a better (shorter) way than the following
hansl function and test cases:
function scalar pd_paneltime(void)
catch pd = nelem(seasonals())
return 1 # assume the case of non-seasonal data
# (but could also be simply no panel-time info)
# artificial example
setobs 5 1:1 --stacked-time-series
setobs 4 1900:1 --panel-time # quarterly
eval pd_paneltime() # gives 4
# real example
open grunfeld # annual
eval pd_paneltime() # gives 1
thanks for any hints
OK, time for some plugging :)
Francesco Valentini and I have just released version 0.9 of the "ketvals"
package, that provides OLS and IV estimation of linear model with
time-varying parameters using a non-parametric technique developed by
Giraitis, Kapetnaios and others.
Download, enjoy, and send us bug reports!
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
1. If there is heterogeneity in the panel model/data, is it true that the
appropriate model is either the Fixed Effect or Random Effects model?
2. I want to choose the appropriate panel regression model when there are
autocorrelations, heterogeneity, multicollinearity, and non-normality. Any
guiding non-highly mathematical materials that you can suggest?
for a simulation study I want to repeat the estimate of a state space
model /N/ times. Repeating the estimate is not a problem, but if I call
the ksmooth function after each estimate (MLE block), gretl allows me it
on the first iteration, but not the following ones.
It appears that the function cannot find the bundle in subsequent iterations
This happens if the repetitions occur in a single script run.
Obviously it doesn't happen if every rep has its own run.
It's really strange!
Is it a bag or am I missing something?
Università del Piemonte Orientale
Dip. di Giurisprudenza e Scienze Politiche,
Economiche e Sociali (DIGSPES)
Questa email è stata esaminata alla ricerca di virus da AVG.
suppose I have a R code which load gretldata via '--send-data' flag and
stores the results RES via 'gretl.export(RES, base)' and 'base' is
$mpirank based string. Is this thread-safe?
I have read a blog stating that if there is heterogeneity in the panel
data the appropriate model is either the Fixed Effect or Random Effects
model, not the OLS Regression. Do you agree with the blogger?
I'm interested to know to determine the appropriate panel regression model
when there are autocorrelations, heterogeneity, multicollinearity, and
non-normality. Appreciate if you suggest a textbook that deals with the
OLS, FE, and RE models for panel data.
Hi gretl users,
there has been an update of the PairPlot package which may be of
interest for some of you. The recent version is 0.8, and it includes a
couple of bug fixes. The changelog is:
Version 0.8 (January 2022):
- Handle eventual missing values in passed series.
- Bugfix of gnuplot's "offset" option: More than at maximum 4
parameters were passed in case of more than 4 features to plot.
- Bugfix: Works now with more than 8 features.
The following commands will install the recent version and show the help
pkg install PairPlot
What are the assumptions to be tested and reported when we use
Fixed-effect and Random-effects models? The same assumptions such as
normality, no autocorrelation, and no heteroscedasticity?
The OP's message below was sent to gretl-users-owner, a backdoor way
of trying to get the attention of gretl people ;-)
I'm forwarding it to the gretl-users list, where it should have been
posted, and where might be of interest to others.
Here's Jack Lucchetti's reply to the original email:
On Tue, 18 Jan 2022, Matt Marcoux wrote:
> Hi all,
> Looking over the documentation for GRETL, and the IRF function will only
> generate an irf after a var/vecm model is estimated.
> However, I am looking to graph the irf for an AR(1) model, Here is my code:
> #open data file
> open ARMAexamples.csv
> #univariate time series AR(1)
> arma 1 0; X --nc
> #matrix ir1 = irf(0,1)
> In stata, we are able to do something similar by using this code. Anyone know
> how one might go about graphing an IRF for an AR(1) model in gretl?
> I am looking to create a graph that looks something like this:
> Thanks for taking the time to read this!
You'll get what you want from the ARMA_IRF function package.
I'll just add: if the term "function package" doesn't resonate with
anyone, please take a look at the "Packages" item under the Help menu
in gretl's main window.
Department of Economics
Wake Forest University