DPANEL and Sargan/Hansen test
by Pindar
Hi,
I'm still trying to get a feeling for the dpanel gmm estimators.
When estimating this xtabond2 statement from Roodman (2006/2008) for
abdata.gdt
"xtabond2 n L.n L(0/1).(w k) yr*, gmmstyle(L.(n w k)) ivstyle(yr*,
equation(level)) robust small"
by
<hansl>
open abdata.gdt
dpanel 1; n const w w(-1) k k(-1) ; \
GMM(n,2,8) GMM(w,2,8) GMM(k,2,8) \
GMMlevel(w,1,1) GMMlevel(k,1,1) --time --sys
<hansl>
I came across a question concerning the Sargan/Hansen test of overid.
restrictions.
In the gretl-guide on p.152 it is stated that "Specifically, xtabond2
computes both a "Sargan
test" and a "Hansen test" for overidentification, but what it calls the
Hansen test is what DPD and
gretl call the Sargan test."
The Hansen test in this example does not reject the validity of the
instruments while the Sargan does.
"Sargan test of overid. restrictions: chi2(100) = 186.90 Prob > chi2 = 0.000
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(100) = 110.70 Prob > chi2 = 0.218
(Robust, but can be weakened by many instruments.)"
In gretl output however the result of the Sargan test is and not the
Hansen test :
"Sargan over-identification test: Chi-square(100) = 154.808 [0.0004]"
That's was quite a surprise for me.
Perhaps it's because the test statistic is not the same and it's really
the Hansen test (cos I believe in what u documented :-)), but why then
such drastic differences?
Cheers
Leon
12 years, 11 months
DPANEL questions concerning accessors, a collapse switch and arguments
by Pindar
Hi,
I think that accessors for the various test-statistics and p-values
would be useful like $ar1, $ar2.
Have you already thought about incorporating a 'collapse' switch?
It would be a good feature for robustness checks.
In my humble opinion the implementation in the (difficult) existing
instrument-matrix-generation-process seems not so hard, since the rows
of the standard matrix just move to the right if there is a zero entry
to the right.
The GMM() function does not support the use of lists right now. Could
this be improved?
Best
Leon
12 years, 11 months
ARIMA models CML
by Alessandro Attanasio
Dear all,
I have a question about ARIMA models. Let us consider, for example, an ARIMA(3,1,3) for the time series y(t). The parameters are estimated using the Conditional Maximum Likelihood, so I use
arima 3 1 3 ; y --conditional
In this case the convergence criterion was not met. I'd like change the convergence tolerance, is it possible?
Thanks.
Best regards
Alessandro
12 years, 11 months
Time varying parameters models
by Logan Kelly
Has anyone worked with time varying parameters models in Gretl? If so, has any one published any example code fragments?
Thanks,
Logan Kelly
13 years
Bootstrapping VAR residuals
by Gabriela Nodari
Dear gretl users,
I would like to obtain simulated data by bootstrapping the residuals of a
VAR model. I have gave a look to the gretl guide, and I found something
related to resampling and bootstrapping (section 7.4), actually, I would
like to adapt the following example - 7.1 in the guide - (with ols) to the
VAR framework:
ols y 0 x
genr ui = $uhat
scalar ybar = mean(y)
scalar replics = 2000
scalar tcount = 0
series ysim = 0
loop replics
ysim = ybar + resample(ui)
ols ysim 0 x
scalar tsim = abs($ coeff (x) / $stderr(x))
tcount +=(tsim > 2.5)
endloop
I have started to write a code.. but my "first" problem is how to save the
matrix of residuals and then generate the new dataset. The next one I
guess will be how to store 2000 Irfs! But let's go step by step..
Would someone have any suggestion?
Thanks in advance.
Gabriela
13 years
Citation: Credit where credit is due.
by Logan Kelly
Hello,
I have been using Gretl for a few months now, and I am ready to start submitting papers written using Gretl. I would like to cite Gretl in those papers. Is there a standard way the developers would like their excellent work cited?
Cheers,
Logan Kelly
13 years
seq function
by Artur Tarassow
Hello,
I just noticed that I cant use an arbitrary positive scalar as the third
argument in the seq function.
This does not work:
matrix sq = seq(0.1,40.2,0.1)
whereas this works:
matrix sq = seq(0.1,40.2,1)
sq
Is there a reason for this limitation? For instance, Gauss' seqa function
accepts any positive scalar.
Artur
13 years
sampling by restriction on a $i variable within a loop
by Luis M. Carrascal
Dear GRTL users,
Probably it is a stupid question, but
how can I restrict data analyses within a loop affecting several response
variables $i, when the restriction affects each variable $i?
the line smpl $i > 0 --restrict does not work within this loop
(from response variable 39 to response variable 145):
loop i=39..145 --progressive
smpl $i > 0 --restrict
quantreg {.95} $i const TMINstd TMINstd2 PRECmedINVstd PRECINVstd2 --robust
more commands
endloop
PD: I have a very large sample size (n=1690)
Thanks you very much in advance
Luis M. Carrascal
<http://www.lmcarrascal.eu/> http://www.lmcarrascal.eu/
email: <mailto:lmcarrascal@mncn.csic.es> lmcarrascal(a)mncn.csic.es
Dept. Biogeografía y Cambio Global
MNCN-Consejo Superior de Investigaciones Científicas
C/ Pinar 25, 3ª planta. 28006 Madrid. SPAIN
Tfno: 34 91 5668963, Fax: 91 5645078
13 years
How to draw a "factorized" histogram?
by yinung＠Gmail
Dear all
I am wondering how to plot a "factorized" histogram (something like
shown below) in similiar to a "X-Y with factor separation" that we
have in gretl now.
Anyone can help?
Thanks...
Yi-Nung Yang
| #
| * #
| # * # *
| # * # *
+____________
1 2
#: group A
*: group B
13 years
sampling by restriction on a $i variable within a loop
by Luis M. Carrascal
Dear Grtl users,
Probably it is a stupid question, but
how can I restrict data analyses within a loop affecting several response
variables $i, when the restriction affects each variable $i?
the line smpl $i > 0 --restrict does not work within this loop
(from response variable 39 to response variable 145):
loop i=39..145 --progressive
smpl $i > 0 --restrict
quantreg {.95} $i const TMINstd TMINstd2 PRECmedINVstd PRECINVstd2 robust
more commands
endloop
PD: I have a very large sample size (n=1690)
Thanks you very much in advance
Luis M. Carrascal
<http://www.lmcarrascal.eu/> http://www.lmcarrascal.eu/
email: <mailto:lmcarrascal@mncn.csic.es> lmcarrascal(a)mncn.csic.es
Dept. Biogeografía y Cambio Global
MNCN-Consejo Superior de Investigaciones Científicas
C/ Pinar 25, 3ª planta. 28006 Madrid. SPAIN
Tfno: 34 91 5668963, Fax: 91 5645078
13 years
