SVAR addon update
by Sven Schreiber
Dear users (or future users) of the SVAR addon,
I guess we forgot to mention the updates to versions 1.51 and 1.52 in
recent months. You should be able to get those from the Help menu under
"Look for addons" (or similar wording in English, sorry I usually run
gretl in the German translation). For new gretl installations these are
automatically included.
The brand new v1.52 is not a big update and essentially includes some
fixes for SVECs (cointegrated SVARs), especially when there are weakly
exogenous variables present. (When the loadings matrix alpha is fixed
and includes some rows with only zeros. - And BTW, "fixes" does not mean
that previous results were wrong, only that something wasn't working
before.)
The previous updates v1.5 and v1.51 released back in September are
perhaps of more general interest; the central new feature is to deal
with heteroscedastic residuals through various new bootstrap variants to
get more reliable confidence bands for the impulse responses in those
cases.
These and other changes are documented inside the PDF help document of
the SVAR addon.
Needless to say, if you find any bugs or anything that isn't working, it
would be helpful to provide feedback, preferably (but not necessarily)
here on this list.
Thanks
sven
6 years
testing for Long Memory using Fractional Integration test
by Yusuf Abduwahab Hassan
Good morning,
Please i want to investigate the integration properties of a variable so as
to analyse the persistence in the series.
1. i want to test the fractional integration hypotheses within the interval
(0,1) with grid ie, with step, 0.1. Is my procedure below correct?
fractint ld_ELCNigeria(0) --all
fractint ld_ELCNigeria(0.1) --all
fractint ld_ELCNigeria(0.2) --all
fractint ld_ELCNigeria(0.3) --all
fractint ld_ELCNigeria(0.4) --all
fractint ld_ELCNigeria(0.5) --all
fractint ld_ELCNigeria(0.6) --all
fractint ld_ELCNigeria(0.7) --all
fractint ld_ELCNigeria(0.8) --all
fractint ld_ELCNigeria(0.9) --all
fractint ld_ELCNigeria(1) --all
2. how can i employ the Gil-Alana(2002) code in Gretl
--
*Yusuf Abdulwahab Hassan.Department of Economics and Development
Studies.Federal University of Kashere,Gombe.+234
8036830166.yabdulwahab(a)fukashere.edu.ng <yabdulwahab(a)fukashere.edu.ng>*
6 years
gretl code news
by Allin Cottrell
I think I forgot to mention a couple of recent developments -- but
sorry if I'm repeating myself -- in response to requests from users.
1) stdize(): there's now a function of this name, documented in git
and snapshots. And, as Ioannis suggested, it has a GUI hook: the
item "Standardize selected series" under the Add menu in the gretl
main window.
2) Number of bootstrap replications for impulse response functions
for VARs/VECMs: the default has been raised from 999 to 1999. And
this number can be changed by the user, either via a new "set"
variable, "boot_iters", or via the GUI dialog to specify details for
IRFs. (As I type this I have the feeling I've already mentioned it,
but oh well.)
Allin
6 years
Panel data unit-root tests
by Victor Olivo
When I try to run unit-root tests with panel-data, I always get the message that it isn’t possible to run the tests because missing data? Is there a solution for this?
Thanks
Sent from my iPhone
6 years
Delete bundle elements within a loop
by Artur Tarassow
Hi all,
I trying to delete bundle elements using a loop. However, this seems to
be forbidden at the moment. Or do I something wrong here? I am using
latest git on Ubuntu 19.10.
<hansl>
bundle b = defbundle("A", 1, "C", 3)
strings S = defarray("A", "B", "C")
loop i=1..nelem(S) -q
string key_name = S[i]
if inbundle(b, key_name)
delete b["@key_name"]
endif
endloop
delete b["A"] # works
</hansl>
Thanks,
Artur
6 years
