nasty bug in gretl 1.7.3
by Allin Cottrell
Hello all,
I've found and fixed a nasty bug lately. Certain gretl commands,
including tobit and some kinds of panel estimation, led to the
creation of a temporary parallel dataset. In 1.7.3, if you have
sub-sampled your main dataset (by some boolean criterion, not just
shifting the sample start or end) and then you carry out one of
these commands, you will not be able to get back to the full
dataset. (The destruction of the temporary dataset confuses
the mechanism for restoring the main dataset.)
I'm looking for your advice here. I'd like to put out a new
release soon. The question is, should this release be based on
current CVS, or should I take the code as of the 1.7.3 release and
patch for this bug (and one or two other small ones)?
If any of you have time, if would be helpful if you can test
current CVS (the Windows snapshot is also up to date) and see if
you can find any breakage. Thanks.
Allin.
16 years, 9 months
help
by Porte Alegre
Hi, I am trying to use gretl. I made forecasts by using Garch model succesfully. But only output options of forecasts I know is console plain text and special databases. Is there any version that processes input and output data over SQL server or common know databases? Best regards,Yuksel
Yeni nesil Windows Live Servisleri’ne şimdi ulaşın! Buraya tıkla!
_________________________________________________________________
Windows Live Messenger'ın için Ücretsiz 30 İfadeyi yükle
http://get.live.com
16 years, 9 months
Re: Re: Re: [Gretl-devel] Questions about the "dataset resample n" command
by andreas.rosenblad＠ltv.se
cottrell(a)wfu.edu @ INTERNET skrev 2008-03-04 15:03:46 :
> On Tue, 4 Mar 2008, andreas.rosenblad(a)ltv.se wrote:
>
> > cottrell(a)wfu.edu @ INTERNET skrev 2008-03-04 03:35:17 :
> > > > On Mon, 3 Mar 2008, andreas.rosenblad(a)ltv.se wrote:
> > >
> > > > After using the command "dataset resample n" to resample >
> > > observations with replacement, how does one restore the original >
> > > full data set?
> > >
> > > You can't. This is not documented yet because I'm not yet sure that
> > > it's useful, or should be retained.
> > > OK. I think the function is very useful and should be retained.
> > It make bootstrapping of all statistical methods implemented in >
> gretl very easy. But to be fully usable there must of course be > a
> command to restore the full data set. I hope it will be > implemented as
well.
>
> The difficulty is that "restoration of the full data set" could not
> mean the same in this case as in the other cases of sampling.
>
> In the other cases, we can "carry back" both modifications to
> existing series that were made when sub-sampled and series that were
> newly created. There's no way to do this with a resampled dataset
> because there's no unique mapping from rows in the modified dataset
> to rows in the original. We could preserve scalars and matrices
> newly modified or created, but we'd have to discard any changes to
series.
>
> With that proviso, it could be done. Allin.
That sounds good.
Best regards
Andreas
16 years, 9 months
matrix error
by Sven Schreiber
The following produces an error on a post-1.7.2-build:
matrix m1 = zeros(1,5)
matrix m2 = ones(1,5)
mymat = m1 | m2
whereas this version works:
matrix m1 = zeros(1,5)
matrix m2 = ones(1,5)
matrix mymat = m1 | m2
I know the second variant (with explicit matrix declaration) is cleaner,
but shouldn't the first one also work?
-sven
16 years, 9 months
Re: Re: [Gretl-devel] Questions about the "dataset resample n" command
by andreas.rosenblad＠ltv.se
cottrell(a)wfu.edu @ INTERNET skrev 2008-03-04 03:35:17 :
> On Mon, 3 Mar 2008, andreas.rosenblad(a)ltv.se wrote:
>
> > After using the command "dataset resample n" to resample >
> observations with replacement, how does one restore the original >
> full data set?
>
> You can't. This is not documented yet because I'm not yet sure that
> it's useful, or should be retained.
OK. I think the function is very useful and should be retained. It make
bootstrapping of all statistical methods implemented in gretl very easy.
But to be fully usable there must of course be a command to restore the
full data set. I hope it will be implemented as well.
Best regards and thanks for a great software,
Andreas
16 years, 9 months
Questions about the "dataset resample n" command
by andreas.rosenblad＠ltv.se
After using the command "dataset resample n" to resample observations with
replacement, how does one restore the original full data set?
Wouldn't it be a better idea to include this dataset resample command in
the "smpl" command syntax instead, so that e.g. "smpl n --resample" gives a
resample of observations with replacements? Then could "smpl full" be used
to restore the full data set. This should be very useful with
ebootstrapping.
Best regards
Andreas
16 years, 9 months
Make "sample" a synonym of "smpl"
by andreas.rosenblad＠ltv.se
Regarding the command "smpl", I find it much more intuitive to write
"sample" instead. Please make "sample" a synonym of "smpl", so that they
both can be used for resetting the sample range.
Best regards
Andreas
16 years, 9 months
Re: Re: [Gretl-devel] Still several outdated screenshots in User's Guide
by andreas.rosenblad＠ltv.se
I have added it to the bug tracker. I have also added a bug that the
command "dataset resample n" is not documented.
Best regards
Andreas
svetosch(a)gmx.net @ INTERNET skrev 2008-02-29 15:57:25 :
> Until somebody feels like contributing the updated screen-shots,
> could you add a bug to the tracker so that it doesn't get forgotten?
>
> (https://sourceforge.net/tracker/?atid=416803&group_id=36234&func=browse)
>
> -> "submit new"
>
> thanks,
> sven
>
> Am 29.02.2008 15:43, andreas.rosenblad(a)ltv.se schrieb:
> > There are still several outdated screenshots in the Gretl User's Guide:
> > > Figure 3.3 and 10.1: Close button still seen
> > Figure 7.1: Labels tab bar missing
> > Figure 16.1: Check buttons for "Show details of iterations" and "Robust
> > standard errors" missing
> > Figure 23.1: Menu entry "Tabular options..." missing
> > > Best regards
> > Andreas
> > > > > > > cottrell(a)wfu.edu @ INTERNET skrev 2008-02-29 14:54:11 :
> > >> This is mostly just a quick fix for correct cooperation with lapack
> >> on 64-bit systems.
> >>
> >> http://gretl.sourceforge.net/index.html
> >>
> >> 2/29/2008 version 1.7.3
> >> - Fix for 64-bit platforms: don't use "long" as the C
> >> counterpart of Fortran 77 "INTEGER"
> >> - Make the rcond function available for asymmetric matrices
> >> - Enable the $yhat accessor for fitted values in equation
> >> systems
> >> - Add Italian function help to Windows build
> >> - Weekly to monthly data compaction: respect the user's
> >> choice of compaction method, as claimed
> >> - Add trimr matrix function (as in Gauss)
> >> - Extend matrix "division" a la Gauss
> >> - Disallow user functions with the same name as a built-in
> >> function
> >> - Update several screenshots in User's Guide
> >>
> >> -- Allin Cottrell
> >> Department of Economics
> >> Wake Forest University, NC
> > > _______________________________________________
> > Gretl-devel mailing list
> > Gretl-devel(a)lists.wfu.edu
> > http://lists.wfu.edu/mailman/listinfo/gretl-devel
> > _______________________________________________
> Gretl-devel mailing list
> Gretl-devel(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-devel
16 years, 9 months
