Gretl project on the launchpad
by Ivan Sopov
Hello, gretl developers.
I'm trying to start a translation of help files into russian on
launchpad.net as it seems to be the most suitable tool to participate
for all familiars with econometrics but not with gettext, linux. cvs,
etc.
The problem is that there is already a project for gretl on launchpad
and it is strongly prohibited to start more than one project for a
single program. I cannot contact with Constantine Tsardounis for about
a month, so I think it is time to re-assign that project to someone
else. On the irc-channel of launchpad I was told that
Our admins can re-assign the project to new owners but we'd prefer to
hear from the upstream owners. can you get one of them to submit a
question here:
https://answers.edge.launchpad.net/launchpad
But if nobody from main developers wants to register and do something
at launchpad it is possible to assign this function to me and in that
case a letter in this list will probably be enough.
I have prepared a .po-file for genr_funcs.xml and gretl_commands.xml
with the help of po4a utility and got 1511 strings for translation
(strings a rather big).
Good luck, Ivan Sopov.
P.S. My previous letter about using launchpad for translation is
http://lists.wfu.edu/pipermail/gretl-devel/2009-November/002171.html
12 years, 9 months
gretl and openmp
by Allin Cottrell
As some of you know, we're currently experimenting with openmp in
gretl. When building from CVS, use of openmp is the default (if
openmp is supported on the host) unless you pass the option
--disable-openmp to the configure script. In addition the current
snapshots for Windows and OS X are built with openmp support
(using gcc 4.4.3 and gcc 4.2.4 respectively).
This note is just to inform you about the state of play, and to
invite submission of test results if people would like to do that.
Right now, we use openmp only for gretl's native matrix
multiplication. So it'll get used (assuming you have at least two
cores) if you do matrix multiplication in a script, or call a
function that does matrix multiplication (such as qform), or use a
built-in command that happens to call matrix multiplication. If we
decide it's a good idea, we could use openmp directives in other
gretl code (but as along as we rely on lapack for much of our
number-crunching, and as long as lapack is not available in a
parallelized form, the scope for threading will remain somewhat
limited).
In a typical current use situation, with gretl running on a
dual-core machine where there's little other demand being placed
on the processors, the asymptotic speed-up from openmp should be
close to a factor of two. However, it takes a big calculation to
get close to the asymptote, and we've found that with small to
moderate sized matrices the overhead from starting and stopping
threads dominates, producing a slowdown relative to serial code.
This is similar to what we found with regard to the ATLAS
optimized blas; see
http://ricardo.ecn.wfu.edu/~cottrell/tmp/gretl_speed.html
Anyway, in case anyone would like to test I'm attaching a matrix
multiplication script that Jack wrote. Right now this is mostly
useful for people building gretl from source, since you want to
run timings both with and without MP, which requires rebuilding.
But if you're currently using a snapshot from before yesterday
(build date 2010-03-21 or earlier) you could run the script, then
download a current snapshot and run it again.
Allin
14 years
smpl in loop
by Marcin Błażejowski
Hi,
two (almost) the same simple scripts:
<script>
nulldata 100
setobs 1 1 --time-series
set echo off
set messages off
series y = normal()
loop i=1..10 --quiet
y[$i] = NA
endloop
smpl --no-missing
printf "valid obs: %d\n", $nobs
</script>
<script>
nulldata 100
setobs 1 1 --time-series
set echo off
set messages off
series y = normal()
loop 1 --quiet
loop i=1..10 --quiet
y[$i] = NA
endloop
smpl --no-missing
printf "valid obs: %d\n", $nobs
endloop
</script>
The only difference is the "loop 1 --quiet" statement. Results of the
first script is 90 (as we expect), but the second is 100.
Is this bug or just limitation of "smpl" command in loops?
Regards,
Marcin
--
Marcin Błażejowski
http://www.wrzosy.nsb.pl/~marcin/
GG# 203127
14 years, 9 months
expressions in loop conditions
by Sven Schreiber
Hi,
this is a feature request for gretl 2.0:
Even simple expressions aren't possible in loop conditions, for example:
"loop for i=1..p-1"
doesn't work, which inflates the trivial LOCs needed in gretl scripts.
Not really serious, I know, but would it be too difficult to make those
things work?
thanks,
sven
14 years, 9 months
devel update
by Allin Cottrell
Hello all,
gretl-devel has been pretty quiet lately, so I thought perhaps I
should put out a status update. It's not that your coders have
been idle lately. On the contrary, Jack and I have had quite a
voluminous correspondence over the past couple of months -- but
the issues we've been discussing have mostly been quite
technical questions that I thought (righly or wrongly) would
probably tax the patience of most gretl-devel subscribers.
Anyway, here are some of the main news points:
* bundles: we have a new data type that hopefully should be very
useful for writers of complex function packages. This is now
written up in chapter 11 of the Gretl User's Guide.
* dynamic panel data models: we have a new command, "hidden" up
to this point, named "dpanel", which represents a substantial
improvement over the old "arbond" command. It can handle
"system GMM" (Blundell-Bond) and also offers greater
flexibility of specification. The one respect in which it's
less flexible than "arbond" is that it does not offer the
option of orthogonal deviations as an alternative to
differencing; it seems to us that this is more of a
theoretical curiosity than a live option of interest to
practitioners. We just have to write up the documentation for
dpanel then we'll expose it for testing.
* Time-series filtering: we now have, I think, a
state-of-the-art implementation of the Butterworth filter
(something that Stephen Pollock talked about at the first
gretl conference in Bilbao last year). You can find it under
the Variable/Filter menu in the GUI when you have loaded a
time-series dataset, or via the bwfilt() function.
* ARIMA: the possibility of handling missing values in the
context of ARIMA estimation has been open ever since we
started using the Kalman filter for "native" exact ML
estimation. We've recently picked this up and, hopefully,
before long we'll be able to offer this facility for all ARIMA
models.
Allin
14 years, 9 months
NA and NaN reloaded
by Sven Schreiber
dear all,
I want to give a little bump/ping to the issue of NA vs. NaN that came
up almost half a year ago; on April 13th Allin wrote (but admittedly
that was not the last message in the thread, so it may not be a conclusion):
<quote>
Good point. I still believe that 0*NA = 0 is fine, so long as NA
(unknown value) and NaN (invalid value) are not conflated. So I
agree we should ensure that the "genr" mechanism does not convert
NaN results to NA.
Strictly, NAs should be confined to values marked as such in
incoming data, values propagated via "genr" on NA input (other
than multiplication by zero), and values marked explicitly by the
user, as in "x[3] = NA" or via "zeromiss()".
</quote>
I'm torn between leaving this issue alone and giving the coders more
time to implement more things that are directly useful, or trying to
argue for this for version 2.0, on the grounds that it yields a
longer-term payoff...
My pragmatic solution would be to enter this into the feature request
tracker (actually it's been a while since I looked there, it may already
be in there... I will run off now to clean up that tracker...)
thanks,
sven
14 years, 9 months
contributing to gretl
by Allin Cottrell
No, I'm not asking for money ;-) But I am asking for something. As
you may know, there are only two regular gretl coders, myself and
Jack Lucchetti. (Others have contributed code from time to time,
for which many thanks.) It's difficult for us to keep up with all
the coding that needs to be done and also to write all the
documentation. I'm thinking that documentation is something that
others could reasonably help with, particularly in the run-up to
gretl 2.0, which is likely to come out later this year.
So here's the suggestion. Please take a look at the Gretl User's
Guide with various questions in mind:
* Are some parts out of date? (Quite likely.) If so, see if you
can update them.
* Are there things missing? The short answer is Yes. For example,
estimation of simultaneous equations systems should have a chapter
in the "Econometric methods" part of the Guide. And I started a
chapter on forecasting but left several sections empty, partly
because I don't have time to work on this and partly because I'm
no expert on the topic. You may notice other omissions.
* Is the organization wrong? (Quite possibly.) Then give us a
suggested reorganization.
If you see something that's out of date or missing, please send us
a draft of an improvement or new text. The User's Guide is in TeX
in its source form, but you don't have to write TeX if that's
unfamiliar to you. And if you're not a native English speaker,
never mind; we can edit what you send.
I'd suggest that if you'd like to write -- or at least make a
start on -- a substantial new section of the Guide, you post to
this list announcing your intention, then we won't have duplicated
effort. When you have something ready to submit, I suggest
emailing it to myself and/or Jack in the first instance, in
whatever format you find easiest.
Allin Cottrell
14 years, 10 months
Fed Database
by Henrique Andrade
Dear Allin,
Is it possible to update the Fed database using Gretl scripts (with no use
of the C programming language)? I know that is possible to download the
entire dataset from the Fed website (
http://research.stlouisfed.org/fred2/downloaddata/), but I would like to
reformat it to the Gretl database format.
Thanks in advance,
--
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge
14 years, 10 months
Strings not translated
by Henrique Andrade
Dear Allin,
Today I've found the following strings that are not translated:
"Expand annual data to quarterly"
"Interpolate higher frequency values"
"Repeat the lower frequency values"
Best regards,
--
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge
14 years, 10 months
new member of the list
by Allin Cottrell
I'd like to welcome Yurii to the gretl-devel list. Yurii contacted
me via the sourceforge messaging system, saying he'd like to join
the gretl team. I suggested he join this list and he has now done
so. Good to have you on board.
Yurii says, "I have extensive experience in data filtering. I have
a proposal to add to Gretl filtering for determine trends." I
think the place to start is with a review of what we offer at
present in gretl with regard to filtering, and what might usefully
be added.
The "shop window" for filtering in gretl is the item
/Variable/Filter in the GUI (when a time-series dataset is
loaded). The other place to look is the Gretl Function Reference,
under the "Filters" heading. Two options have been added since the
last release (1.9.1), namely the Butterworth filter and polynomial
trend fitting, so to get up to date it's necessary to check
current CVS (and/or the binary snapshots).
The other place to look for discussion of filtering options in
gretl is section 5.3 of the Gretl User's Guide, "Time-series
filters." This has also been updated recently, to include an
account of the Butterworth filter.
Allin
14 years, 10 months
