1:50 p.m.
Hello,
it's been a while since I compiled the cvs version, so today was the
first time I noticed the "you can also build the pdf docs" part of the
configure output (on Ubuntu 8.04). I tried that but got errors
complaining about mathtime.sty missing. I'm using the standard Texlive
distro packaged in/for ubuntu, no special stuff AFAIK.
For some background on mathtime in texlive see
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=425460.
Could the dependency on the mathtime package be removed? I'm using this
latex installation for lots of math without errors, so my first guess
would be that it's possible.
thanks,
sven
BTW, the list has been quiet recently, is that good or bad? Oh, and
another thing, I wanted to log in to the gretl area at sourceforge, but
couldn't get a secure https connection, only http. Has anybody else
noticed that?
2:15 p.m.
On Fri, 7 Nov 2008, Sven Schreiber wrote:
it's been a while since I compiled the cvs version, so today was
the first time I noticed the "you can also build the pdf docs"
part of the configure output (on Ubuntu 8.04). I tried that but
got errors complaining about mathtime.sty missing. I'm using the
standard Texlive distro packaged in/for ubuntu, no special stuff
AFAIK.
There's no reference to mathtime in the current TeX manual
sources: if the --with-lucida option is given we use lucidabr,
otherwise we use the regular TeX fonts, with the "ae" package to
get T1 encoding. Somehow you must have some stale files. Try rm
-f *a4.tex in doc/tex. And be sure to ./configure.
BTW, the list has been quiet recently, is that good or bad?
Hmm, hard to tell. I've been very busy with other things, so
I'm not too upset at having relatively few postings to respond to
right now.
Oh, and another thing, I wanted to log in to the gretl area at
sourceforge, but couldn't get a secure https connection, only
http. Has anybody else noticed that?
Quite a lot has changed at sf with their move to their new Chicago
data center. See
http://sourceforge.net/community/forum/forum.php?id=11&page
Allin.
2:22 p.m.
Allin Cottrell schrieb:
On Fri, 7 Nov 2008, Sven Schreiber wrote:
> it's been a while since I compiled the cvs version, so today was
> the first time I noticed the "you can also build the pdf docs"
> part of the configure output (on Ubuntu 8.04). I tried that but
> got errors complaining about mathtime.sty missing. I'm using the
> standard Texlive distro packaged in/for ubuntu, no special stuff
> AFAIK.
There's no reference to mathtime in the current TeX manual
sources: if the --with-lucida option is given we use lucidabr,
otherwise we use the regular TeX fonts, with the "ae" package to
get T1 encoding. Somehow you must have some stale files. Try rm
-f *a4.tex in doc/tex. And be sure to ./configure.
This is a fresh cvs checkout on a new system where gretl has never been
compiled before, so I'm 99% sure there cannot be stale files in the
build directory.
And if I only do ./configure then I don't get the option to build the
docs. So I had done ./configure --enable-build-doc before, with the
stated error result.
Are you sure the regular Tex fonts are used, I thought I remembered that
Times was used as a substitute for Lucida. Or is that what you meant?
thanks,
sven
2:40 p.m.
On Fri, 7 Nov 2008, Allin Cottrell wrote:
On Fri, 7 Nov 2008, Sven Schreiber wrote:
> it's been a while since I compiled the cvs version, so today was
> the first time I noticed the "you can also build the pdf docs"
> part of the configure output (on Ubuntu 8.04). I tried that but
> got errors complaining about mathtime.sty missing. I'm using the
> standard Texlive distro packaged in/for ubuntu, no special stuff
> AFAIK.
There's no reference to mathtime in the current TeX manual
sources: if the --with-lucida option is given we use lucidabr,
otherwise we use the regular TeX fonts, with the "ae" package to
get T1 encoding. Somehow you must have some stale files. Try rm
-f *a4.tex in doc/tex. And be sure to ./configure.
Actually, mathtime is used in the Portuguese translation.
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
2:54 p.m.
Allin Cottrell schrieb:
On Fri, 7 Nov 2008, Riccardo (Jack) Lucchetti wrote:
> On Fri, 7 Nov 2008, Allin Cottrell wrote:
>> There's no reference to mathtime in the current TeX manual
>> sources...
> Actually, mathtime is used in the Portuguese translation.
Well spotted. I've now removed that.
thanks for the quick fix, it's working fine now.
-sven
4:36 p.m.
Sorry about that one.
Must be some very old include.
Hélio
On Fri, Nov 7, 2008 at 2:54 PM, Sven Schreiber <svetosch(a)gmx.net> wrote:
Allin Cottrell schrieb:
> On Fri, 7 Nov 2008, Riccardo (Jack) Lucchetti wrote:
>
>> On Fri, 7 Nov 2008, Allin Cottrell wrote:
>>> There's no reference to mathtime in the current TeX manual
>>> sources...
>> Actually, mathtime is used in the Portuguese translation.
>
> Well spotted. I've now removed that.
>
thanks for the quick fix, it's working fine now.
-sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
4:45 p.m.
New subject: any documentation on error message
Hi,
I got below error message when I tried to calculate mean on a matrix (use meanc).
"Data types not conformable for operation"
Does anyone know what is reason? Also is there any documentation I can find the detailed
help for such error messages?
Thank you!
Regards,
Rebecca
5:11 p.m.
New subject: any documentation on error message
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
I got below error message when I tried to calculate mean on a
matrix (use meanc).
"Data types not conformable for operation"
Does anyone know what is reason?
Can you show the command in context? This function take a matrix
as argument and returns a row vector, as in
matrix M = mnormal(4,2)
matrix v = meanc(M)
print M v
If I try, e.g.,
scalar s = meanc(M)
I get the error message you mentioned (unless M has just one
column, in which case gretl will convert the result to a scalar).
You can never assign the result of meanc() to a variable of
series type.
Also is there any documentation I can find the detailed help for
such error messages?
You really just have to cross-reference these errors with the
"function reference" to see why they arise.
Allin Cottrell
6:44 p.m.
New subject: any documentation on error message
Thank you, Allin.
I have a further question.
I have time series data for 50 variables. I want to calculated the mean for each variable
and store the data as a cross-section dataset that has two columns: one for 50 means and
one for their correpeponding varibles.
Could you tell me how to do that in gretl? Thank you!
Regards,
Rebecca
--- On Fri, 11/7/08, Allin Cottrell <cottrell(a)wfu.edu> wrote:
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] any documentation on error message
To: zzh5460(a)yahoo.com, "Gretl list" <gretl-users(a)lists.wfu.edu>
Date: Friday, November 7, 2008, 9:11 AM
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
> I got below error message when I tried to calculate
mean on a
> matrix (use meanc).
>
> "Data types not conformable for operation"
>
> Does anyone know what is reason?
Can you show the command in context? This function take a
matrix
as argument and returns a row vector, as in
matrix M = mnormal(4,2)
matrix v = meanc(M)
print M v
If I try, e.g.,
scalar s = meanc(M)
I get the error message you mentioned (unless M has just
one
column, in which case gretl will convert the result to a
scalar).
You can never assign the result of meanc() to a variable of
series type.
> Also is there any documentation I can find the
detailed help for
> such error messages?
You really just have to cross-reference these errors with
the
"function reference" to see why they arise.
Allin Cottrell
7:06 p.m.
New subject: any documentation on error message
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
I have time series data for 50 variables. I want to calculated
the mean for each variable and store the data as a cross-section
dataset that has two columns: one for 50 means and one for their
correpeponding varibles.
Sorry, I don't understand: the first (new) column is to contain 50
means, but what is the second column to contain?
You can create a series that contains the means of the variables
in the current dataset (provided the number of variables is not
greater than the number of observations) using something like:
list L = dataset
series means = 0
loop foreach i L
means[i] = mean($i)
endloop
--
Allin Cottrell
Department of Economics
Wake Forest University
4:37 p.m.
New subject: any documentation on error message
I've been using gretl for a recent project and I have a couple of
quesions/observations about syntax.
In Allin's code snippet below, why would you use means[i] instead of
means[$i] in the loop. I've used
loop i=start..end
$i
endloop
which seems to work with one letter loop variables, i, j, k, etc., but not
loop iloop=start..end
$iloop
endloop
when "iloop" is not recognized. With that I had to construct a
scalar iloop=start-1
loop while iloop < end
iloop=iloop+1
iloop
endloop
Also, my current script is a couple of hundred lines long with a couple
of big loops. If I run the script and there's an error, sometimes I
cannot "recover" the script because it's stopped in the middle and left
something "hanging" so that if I correct the error and re-run, the
script doesn't actually run, the output is simply a list of the script.
I just kill the session and re-start, which works. Is there a different
method for killing the script. A couple of times I highlighted "endloop"
and ran just that, which caused gretl to crash.
I also noticed that if X is a matrix
X[i,j]^(exponent)
does not work, when
temp=X[i,j]
temp^(exponent)
is necessary.
Finally, my data set consists of 15146 cross-sectional units with
varying observations over time. Monthly data within the interval 1962:01
through 2007:12. Any way to get gretl to recognize a pattern like this
as some sort of panel?
Thanks.
Rob
Allin Cottrell wrote:
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
> I have time series data for 50 variables. I want to calculated
> the mean for each variable and store the data as a cross-section
> dataset that has two columns: one for 50 means and one for their
> correpeponding varibles.
Sorry, I don't understand: the first (new) column is to contain 50
means, but what is the second column to contain?
You can create a series that contains the means of the variables
in the current dataset (provided the number of variables is not
greater than the number of observations) using something like:
list L = dataset
series means = 0
loop foreach i L
means[i] = mean($i)
endloop
8:28 p.m.
New subject: any documentation on error message
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
Thank you, Allin.
I have a further question.
I have time series data for 50 variables. I want to calculated the mean
for each variable and store the data as a cross-section dataset that has
two columns: one for 50 means and one for their correpeponding varibles.
Could you tell me how to do that in gretl? Thank you!
Is this what you mean?
<script>
set echo off
set messages off
nulldata 50
a = normal()
b = normal()
c = normal()
list X = a b c
outfile @dotdir/foo.csv --write
printf ",mean\n"
loop foreach i X -q
printf "$i,%12.5f\n", mean($i)
end loop
outfile @dotdir/foo.csv --close
open @dotdir/foo.csv
</script>
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
8:59 p.m.
New subject: any documentation on error message
Thank you, Jack.
Here is what I mean:
I have time series (daily) data for about 50 portfolios. I want to calculate the mean over
time for each of the portfolios. And store them in below way:
portfolio 1 mean 1
portfolio 2 mean 2
portfolio 3 mean 3
...
I am able to get statics summary for all the portfolios. (That includes mean and std) How
to store these statics for each portfolio in a way convenient to further processing?
Thank you!
Regards,
Rebecca
--- On Fri, 11/7/08, Riccardo (Jack) Lucchetti <r.lucchetti(a)univpm.it> wrote:
From: Riccardo (Jack) Lucchetti <r.lucchetti(a)univpm.it>
Subject: Re: [Gretl-users] any documentation on error message
To: zzh5460(a)yahoo.com, "Gretl list" <gretl-users(a)lists.wfu.edu>
Date: Friday, November 7, 2008, 12:28 PM
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
> Thank you, Allin.
>
> I have a further question.
>
> I have time series data for 50 variables. I want to
calculated the mean for each variable and store the data as
a cross-section dataset that has two columns: one for 50
means and one for their correpeponding varibles.
>
> Could you tell me how to do that in gretl? Thank you!
Is this what you mean?
<script>
set echo off
set messages off
nulldata 50
a = normal()
b = normal()
c = normal()
list X = a b c
outfile @dotdir/foo.csv --write
printf ",mean\n"
loop foreach i X -q
printf "$i,%12.5f\n", mean($i)
end loop
outfile @dotdir/foo.csv --close
open @dotdir/foo.csv
</script>
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
2:15 a.m.
New subject: any documentation on error message
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
Here is what I mean:
I have time series (daily) data for about 50 portfolios. I want to calculate the mean
over time for each of the portfolios. And store them in below way:
portfolio 1 mean 1
portfolio 2 mean 2
portfolio 3 mean 3
...
I am able to get statics summary for all the portfolios. (That
includes mean and std) How to store these statics for each
portfolio in a way convenient to further processing?
I think that Jack and I have told you that: how to turn a list of
statistics into a data series. Take either of our procedures and
try it.
Allin Cottrell
9:23 a.m.
gretl-users-bounces(a)lists.wfu.edu skrev 2008-11-07 14:50:51 :
Hello,
[snip]
thanks,
sven
BTW, the list has been quiet recently, is that good or bad?
I am busy learning C so I can join Jack and Allin in the development of
gretl. I hope to help expand it to a fully featured statistical package. My
first goal will be to include all the 100 statistical tests from here:
http://www.sagepub.com/booksProdDesc.nav?currTree=Courses&level1=Cour...
in gretl.
Andreas
5647
days inactive
5653
days old
15 comments
7 participants
participants (7)
-
Allin Cottrell -
andreas.rosenblad＠ltv.se -
Hélio Guilherme -
Rebecca Zhang -
Riccardo (Jack) Lucchetti -
Rob Tischer -
Sven Schreiber