Stefano Balietti | 1 Jun 23:54 2008
Picon

Suppress warnings

Can I suppress Gretl warnings about numerical computation somehow? I
mean msgs like:

'^': Numerical argument out of domain: missing values were generated

exp(): Numerical result out of range: missing values were generated

That is because within an mle block i can get several warnings, but it
is basically impossible to avoid it.

Cheers
Talha Yalta | 2 Jun 07:23 2008
Picon

Re: Suppress warnings

When you get a warning like that, you are supposed to do something
about it not ignore it. Do you want to get the correct result or some
result correct or not?
Regards
ATY

On Mon, Jun 2, 2008 at 12:54 AM, Stefano Balietti
<futur.dorko@...> wrote:
> Can I suppress Gretl warnings about numerical computation somehow? I
> mean msgs like:
>
> '^': Numerical argument out of domain: missing values were generated
>
> exp(): Numerical result out of range: missing values were generated
>
> That is because within an mle block i can get several warnings, but it
> is basically impossible to avoid it.
>
> Cheers
> _______________________________________________
> Gretl-users mailing list
> Gretl-users@...
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>

--

-- 
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--
(Continue reading)

Stefano Balietti | 2 Jun 10:24 2008
Picon

Re: Suppress warnings

On Mon, Jun 2, 2008 at 7:23 AM, Talha Yalta <talhayalta@...> wrote:
> Do you want to get the correct result or some
> result correct or not?

Since warnings do not necessarily mean errors, and as my results are
correct finally, I would like to ignore them.

Regards
Riccardo (Jack) Lucchetti | 2 Jun 10:57 2008
Picon

Re: Suppress warnings

On Mon, 2 Jun 2008, Talha Yalta wrote:

> When you get a warning like that, you are supposed to do something
> about it not ignore it. Do you want to get the correct result or some
> result correct or not?

It may well happen, during an iterative procedure (like mle for example) 
that an attempt is made to compute, say, the log of a negative number. In 
these cases, the warning could be transient, if in the end the algorithm 
converges.

Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche

r.lucchetti@...
http://www.econ.univpm.it/lucchetti
On Mon, 2 Jun 2008, Talha Yalta wrote:

> When you get a warning like that, you are supposed to do something
> about it not ignore it. Do you want to get the correct result or some
> result correct or not?

It may well happen, during an iterative procedure (like mle for example) 
that an attempt is made to compute, say, the log of a negative number. In 
these cases, the warning could be transient, if in the end the algorithm 
converges.

Riccardo (Jack) Lucchetti
(Continue reading)

Allin Cottrell | 2 Jun 17:57 2008
Picon

Re: Suppress warnings

On Sun, 1 Jun 2008, Stefano Balietti wrote:

> Can I suppress Gretl warnings about numerical computation somehow? I
> mean msgs like:
> 
> '^': Numerical argument out of domain: missing values were generated
> 
> exp(): Numerical result out of range: missing values were generated
> 
> That is because within an mle block i can get several warnings, but it
> is basically impossible to avoid it.

Try "set warnings off" in current CVS.

I'm not sure this will work for mle, because I can't find an 
example to hand that generates such a warning.  It does work
for, e.g.,

nulldata 20
set warnings off
genr x = normal() + .8
genr lx = log(x)

Allin.
Louis Luangkesorn | 4 Jun 02:34 2008
Picon

Bug report: Ramanathan ps2_1.inp

This is for 

/gretl/share/scripts/ps2-1.inp

the practice script file for Ramanathan.  Bug is current for file version 1.1.1.1 in CVS

The comments in this file are of the form '(*'  '*)' as opposed to '#'   causing syntax errors.  In addition, the
line 'open data2-1' is omitted.  This file should be:

-- ps2-1.inp begins --
open data2-1
# PS2.1 using DATA2-1, illustrating frequency distributions, Section 2.
help freq
# Because SHAZAM and gretl have different command structures, the outputs don't correspond exactly *)
freq vsat
freq msat
-- ps2-1.inp ends --

Louis

"Do one thing every day that scares you." – Eleanor Roosevelt

Allin Cottrell | 4 Jun 05:40 2008
Picon

Re: Bug report: Ramanathan ps2_1.inp

On Tue, 3 Jun 2008, Louis Luangkesorn wrote:

> This is for 
> 
> /gretl/share/scripts/ps2-1.inp
> 
> the practice script file for Ramanathan.  Bug is current for 
> file version 1.1.1.1 in CVS
> 
> The comments in this file are of the form '(*' '*)' as opposed 
> to '#' causing syntax errors.  In addition, the line 'open 
> data2-1' is omitted.  This file should be...

Thanks, you're right.  Somehow that file was missed when the other 
ones were modernized.

Allin Cottrell
Wilson Mixon | 6 Jun 01:15 2008

More on systems

Two observations/questions about systems estimation via the dialog box:
1. Tests are not recorded in the command log.
2. Could coefficient sums be added to the test menu list?
Sven Schreiber | 6 Jun 10:32 2008
Picon
Picon

Re: More on systems

Wilson Mixon schrieb:

> 2. Could coefficient sums be added to the test menu list?

Personally I don't think this is useful enough to be added to a menu. Or 
can you explain why or when such a hypothesis is regularly tested?

cheers,
sven
Thomas La Bone | 7 Jun 20:19 2008

Debian Repository for Latest Version of Gretl


Can someone tell me the repository for the latest debian version of 
Gretl? Specifically, I am looking for the name of the repository as it 
would be entered into synaptic, e.g., deb http:// blah blah blah (with 
the blahs replaced with something meaningful). Thanks.

Tom


Gmane