Fwd: Orthonormal and orthogonal
by Sven Schreiber
I'm forwarding a question from Timmy to the list. See below
cheers
sven
-------- Weitergeleitete Nachricht --------
Betreff: Orthonormal and orthogonal
Datum: Thu, 14 Mar 2019 16:13:33 +0100
Von: Olasehinde Timmy <timmexdareal(a)gmail.com>
An: Sven Schreiber <svetosch(a)gmx.net>
Dear Prof,
I should have sent you this message via the gretl mailing list, but it
seems there is a technical problem with the server. I would be glad if
you can show me how to compute orthonormal and orthogonal of a matrix in
gretl. For example, if alpha matrix is α = {-0.2,-0.3;0.1,-0.4}.
Thanks for your help.
Regards
Timmy.
6 years, 11 months
random-effects probit problem
by Allin Cottrell
A heads-up: I'm afraid the random-effects probit "plugin" (loabable
module) got broken in the gretl 2019a release for Mac OS X. Thanks
to Jeff Bowman for pointing this out. The problem is fixed in the
current gretl snapshot for Mac.
The problem is obvious if you try to use random-effects probit: you
get an error message saying the plugin can't be loaded. The reason
for this is that the plugin was still linked against the previous,
out-of-date, version of libgretl. Only the one plugin is affected.
Note to self: be sure to do "make clean" before building gretl for a
release!
Allin Cottrell
6 years, 12 months
dbnomics: problem retrieving data bundle
by klaus.hasenbach＠web.de
Hi,
Untill mid of january everything worked very good.
But now I have problems retrieving data by bundl (windows 10, Gretl
2018c or 2019a)
1.)Doing it for OECD data it is very slow (what was done within a minute
is done now in many long minutes)
2.) Doing it for BIS it does not work at all, error messages occurre:
/? list X = dbnomics_bundles_to_list( bs, "series_code" )//
//Kein Datensatz aktiv//
//
//Fehler bei Skriptausführung: Stopp//
//> list X = dbnomics_bundles_to_list( bs, "series_code" )/
Below I show a script examble for OECD and BIS:
open dbnomics
set verbose on
include dbnomics.gfn
nulldata 870
setobs 12 1948.1
provider = "OECD"
database = "KEI"
bundle spec = defbundle("mask",".G-20+DEU+USA..M")
bs = dbnomics_get_multiple(provider, database, 1000, 0, spec)
dbnomics_bundles_print(bs)
list X = dbnomics_bundles_to_list( bs, "series_code" )
printf "\nHere are the series in list X:\n"
list X print
provider = "BIS"
database = "EERI"
bundle spec = defbundle("mask","M+Q.R.B.")
bs = dbnomics_get_multiple(provider, database, 1000, 0, spec)
dbnomics_bundles_print(bs)
list X = dbnomics_bundles_to_list( bs, "series_code" )
printf "\nHere are the series in list X:\n"
list X print
Any help. Thank you.
Klaus
7 years
RVNR.gfn package to be retired
by Sven Schreiber
Dear all,
this is a warning notice that the contributed function package RVNR.gfn
by Guillermo Verduzco Bustos will be removed from the function package
server soon. The package has certain technical issues and the author
could not be contacted for quite a while now. It hasn't been updated
since 2015.
(If you are --or know-- Guillermo, you are most welcome to contact me
on- or off-list.)
If you need this package for your work, you would have to keep a local
copy of it.
Generally speaking the package provides a nonparametric test for
autocorrelation in a time series. Some other nonparametric tests for
autocorrelation are available in the FEP.zip package. (This package is
by Artur Tarassow and myself, so I'm sorry this must sound like an ad,
but this is what came to my mind as related functionality.)
Have a good weekend everybody!
Sven
7 years
gretl Function on Managed Chromebooks
by Csakai, Damian
Good Morning,
We have a school district inquiring about the use of gretl on Chromebooks. The Chromebooks throughout the schools are managed. Does the gretl software require Windows/MAC OSX? Is there an application that can be pushed to Chromebooks that are managed? Is there a web URL that can be used on a Chromebook if the gretl software can be installed/hosted on a local server?
Thank you,
Damian Csakai
Customer Engineer - II
mindSHIFT, a Ricoh Company
RICOH USA, INC
711 Third Avenue, Suite 205
New York, NY 10017
Office: (212)253-3509
Damian.Csakai(a)mindshift.com
NOTICE: This email
with its attachments, if any, is intended solely for permitted use by
authorized recipients. If you are not an authorized recipient of this email,
your retention, transmission, disclosure, or use of this email is prohibited.
If you received this email in error, please notify the sender and delete this
email and its attachments from your devices and systems.
7 years
DFM package question
by ΑΝΔΡΕΑΣ ΖΕΡΒΑΣ
Hi all,
I have a question regarding DFM package, version 0.2 on Gretl 2018c (or
b) . It is about using unbalanced time series. I understand that dynamic
factor models are estimated on the common sample of the time series.
However, my understanding of the literature is that if some of these
time series from which the factors are extracted have longer sample,
then it is possible to use the Kalman filter to fill the missing values
of the dynamic factors. In the help such a feature is not mentioned. Is
it possible to do it in a quick way, or we would have to wait for a new
version?
Many thanks.
Andreas
7 years
DFM package question
by ΑΝΔΡΕΑΣ ΖΕΡΒΑΣ
The possibility to treat missing observations is mentioned explicitly in
Giannone, Reichlin and Sala JME (2008) for the case of forecasting the
factors when some of the time series from which they are extracted are
missing in the context of nowcasting GDP using a DFM model. I would also
be interested to know if this is also possible to be done backwards,
that is to extract the factors in the part of the sample where some of
the time series are missing, and how (I guess the way would be the same
no matter where one goes, forward or backwards).
Many thanks again.
Andreas
7 years
The "fcast" command and panel data
by Allin Cottrell
This is a follow-up to Artur's comment/question in
http://lists.wfu.edu/pipermail/gretl-users/2019-February/013703.html
Use of "fcast" with panel data is now enabled (to varying degrees)
for pooled OLS, fixed effects, random effects and dpanel. By "now" I
mean in git and the snapshots.
Preliminary note: if you want to produce out-of-sample predictions
using panel data you should ensure that any required transformations
(lags, differences, etc.) are generated for the entire dataset
before a model is estimated on a sub-sample.
Within sample forecasts
Producing within-sample "forecasts" is unproblematic, if in general
not very interesting since the $yhat accessor for fitted values is
already available. However, there's modest added value from "fcast"
in a couple of respects: you can get a printout of the forecast
evaluation statistics; and for random-effects estimation the values
produced by fcast include an estimate of the individual effect,
which is omitted from $yhat.
Out of sample forecasts
In the panel data context "out of sample" could mean either in the
time dimension or in the cross-sectional dimension. As of now,
gretl's automatic --out-of-sample option operates strictly in the
time dimension, but it's possible to forecast out of sample in the
cross-sectional dimension via the stobs/endobs arguments to "fcast"
or by setting the sample range appropriately before calling "fcast".
I'll deal with the --out-of-sample option (time dimension) first.
This is subject to one restriction and one caveat.
Restriction: the model on which forecasts are to be based must be
estimated using all individuals in the dataset (i.e. sub-sampled
only with respect to time). This is not a very severe limitation:
one can always save, then open, a reduced copy of the dataset
(sub-sampled in the cross-sectional dimension) first, if necessary.
Caveat: It's not clear what we should do if the model specification
includes time dummies. At present we impute an out-of-sample time
effect equal to the mean of the in-sample time effects. This is
obviously debatable; feel free to debate it. Perhaps we should ban
such prediction altogether.
In addition, a refinement: for models estimated via "dpanel", the
--out-of-sample option produces a dynamic forecast by default (as
with dynamic estimators on straight time-series data). This can be
interdicted by use of the --static option to "fcast".
Now to out-of-sample in the cross-sectional dimension. What are we
going to do with individual effects, for the fixed- and
random-effects estimators? For now, we set the predicted individual
effect to the global constant (for fixed effects) or to its expected
value of zero (for random effects). This is obviously another
debatable point, analogous to the out-of-sample time effect case.
Besides debatable policies there are no doubt some bugs lurking in
the new fcast facilities. Comments welcome.
Allin
7 years
"setobs --panel-time" option doesn't work
by Artur Tarassow
Hi,
I've just stumbled about the --panel-time option for setobs. One of the
things that always annoyed me was the lack of this feature. I am really
happy its available now!
However, the following doesn't work for me using latest git on ubuntu:
<hansl>
set verbose off
clear
N = 2 # cross-sectional dim.
T = 14 # time dim.
scalar NT = N*T
nulldata NT --preserve
setobs T 1:1 --stacked-time-series
#setobs 7 2019-02-26 --panel-time
setobs 4 1990:1 --panel-time
series x = normal()
series y = normal()
print y x -o
</hansl>
However, the printout of the dataset does not comprise any information
on dates and frequencies:
<output>
y x
1:01 -1.018049 3.117621
1:02 -1.139177 -1.372563
1:03 0.304792 1.633629
1:04 -0.837392 0.612071
1:05 -1.555112 -1.502532
1:06 0.165970 0.160983
1:07 1.287358 2.031679
.
.
.
.
</output>
Best,
Artur
7 years
