Sorry if this is too off-topic, as we may not implement this in R
(although we may do so).
A student of mine is looking at some public opinion data in which there
appears to be a statistically significant difference between levels of
support for a proposal based on which of two question wording
Yes, typically one uses lme4 or nlme to do multilevel modeling in R.
--
Andrew J Perrin - andrew_perrin (at) unc.edu - http://perrin.socsci.unc.edu
Assistant Professor of Sociology; Book Review Editor, _Social Forces_
University o
Shouldn't that be
kappa = rater1 - rater2 / (1-chance)
?
Andy
--
Andrew J Perrin - andrew_perrin (at) unc.edu - http://perrin.socsci.unc.edu
Assistant Professor of Sociology; Book Review Editor, _Social Forces_
University of No
I can speak to some of these issues. I don't know about how much benefit
you can get from SMP for *single* instances of R, though.
1.) Multicore will be helpful, at least, if you are running several
instances of R at once. So, for example, if you have people running two
different models at the
University of North Carolina - CB#3210, Chapel Hill, NC 27599-3210 USA
New Book: http://www.press.uchicago.edu/cgi-bin/hfs.cgi/00/178592.ctl
On Thu, 8 Mar 2007, Uwe Ligges wrote:
>
>
> Andrew Perrin wrote:
>> Greetings-
>>
>> Running R 2.4.0 under Debian Linux, I am getti
Greetings-
Running R 2.4.0 under Debian Linux, I am getting a memory error trying to
read a very large file:
> library(foreign)
> oldgrades.df <- read.spss('Individual grades with AI (Nov 7
> 2006).sav',to.data.frame=TRUE)
Error: cannot allocate vector of size 10826 Kb
This file is, granted,
It's BibTeX source -- used for the BibTeX bibliography management system
that integrates with LaTeX.
http://www.ecst.csuchico.edu/~jacobsd/bib/formats/bibtex.html
http://www.ctan.org
--
Andrew J Perrin - andrew_perrin (at) unc.e
At step 5, don't copy the file. Instead, do:
make install
--
Andrew J Perrin - andrew_perrin (at) unc.edu - http://perrin.socsci.unc.edu
Assistant Professor of Sociology; Book Review Editor, _Social Forces_
University of North Ca
Greetings - is there any update on a PostgreSQL driver for the DBI
package? If not, what's the currently-preferred method of creating a link
from a PostgreSQL database and R?
Thanks.
--
Andrew J Perrin - andrew_perrin (at) unc.
What, in particular, do you want to do? Content analysis is a mode of
research, not a statistical technique.
--
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociology, U of North Carolina, Chapel Hill
[EMAI
On Thu, 27 May 2004, Greg Tarpinian wrote:
> I apologize if this has been addressed before;
> recently
> I read an article in Forbes which discussed how SCO
> was going after companies that have been using Linux.
> The article made the point that the ideas behind GPL
> are under attack precisely b
Brian Ripley wrote:
>I can't see this in your examples: which viewer did you use? I do see it
>in phantom.ps.
>
>However, looking at resources.bygt.eps, the file is corrupt with a load
>of bytes after
>
>%%EOF
>
>starting with nulls and then some postscript. So I don't think this is an
>R problem
Greetings-
An odd situation has developed. I use the following code to create .eps
files of two very similar graphs:
postscript(file='resources.bygt.eps', onefile=FALSE, horizontal=TRUE)
barplot(resources.bygt.matrix,
beside = TRUE,
legend.text=c('narrative','doubt'),
name
Generally, use natbib style and then
\citep-> (Anderson 1992)
\citet-> Anderson (1992)
\citeyearpar -> (1992)
\citealt -> Anderson 1992
--
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociology, U
I have been using a little function I wrote myself; look at
http://www.unc.edu/home/aperrin/tips/src/icc.R for the code. Not pretty,
but it works.
ap
--
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociol
On Fri, 21 Nov 2003, Douglas Bates wrote:
> Andrew Perrin <[EMAIL PROTECTED]> writes:
>
> > Sorry for my ignorance, but could you explain a little further? I'm
> > guessing from your response that this makes the log-likelihood that is
> > quoted by glmmPQL a p
ood, and what is being quoted is
> not a likelihood for the original problem.
>
> On Fri, 21 Nov 2003, Andrew Perrin wrote:
>
> > Greetings-
> >
> > a reviewer for a paper of mine noted an anomaly in some models I ran using
> > glmmPQL (from the MASS package
Greetings-
a reviewer for a paper of mine noted an anomaly in some models I ran using
glmmPQL (from the MASS package). Specifically, the models are three-level
hierarchical probit models estimated using PQL under R. The anomaly is
that the log-likelihoods decrease (or, alternatively -2logLik inc
For continuous, see the lme package. For categorical, see glmmPQL in the
MASS package, or GLMM in LME4.
ap
--
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociology, U of North Carolina, Chapel Hill
[EMAIL
Three suggestions:
1.) Raudenbush and Bryk, _Hierarchical Linear Models: Second Edition_
(Sage, 2002)
2.) Pinheiro and Bates, _Mixed-Effects Models in S and S-Plus_ (Springer)
3.) Fox, _An R and S-Plus Companion to Applied Regression_, plus the
appendix available via the web on multilevel models
[EMAIL PROTECTED],
[EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED],
[EMAIL PROTECTED]
Subject: [ibiblio.org #1674] Mirroring request
Resent-Date: Tue, 1 Jul 2003 19:34:05 -0400 (EDT)
Resent-From: Andrew Perrin <[EMAIL PROTECTED]>
Resent-To: [EMAIL PROTECTED]
Resent-Subject:
This is untested, so try it first, but my approach would be:
W<- function(w)
{
W[i]<-ifelse(comp[i,1]==1, Wo, (work[i-1,5]work[i-1,13]*/Ef))
}
Best,
Andy Perrin
--
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professo
> check up ?mtext()
Thanks, but that's not what I'm looking for. I don't want to change the
text string but the axis labels themselves.
ap
--
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociology, U of N
Please forgive my ignorance on grapics. I'm trying to make a relatively
simple plot with two line plots, same axes, by mean over a series of
dates. I can make the plot well like this:
plot(sort(tapply(first.anti.auth.sum,date,mean), partial=1), type="l",
col="yellow",ylim=c(0,2.0))
par(new=TRUE)
p
I am foggy on this myself, but I *think* it is inferred from the grouping
structure in the call to (n)lme or in the groupedData data structure. Have
a look at ?groupedData in R for more details.
ap
--
Andrew J Perrin - http://www
On Tue, 24 Jun 2003, Dirk Eddelbuettel wrote:
>
> On Tue, Jun 24, 2003 at 07:47:19PM -0400, Andrew Perrin wrote:
> > OK, thanks. Is that the problem with e1071? Any advice on getting it
> > loaded?
>
> [ It is considered impolite to reply to private messages via th
] * andrew_perrin (at) unc.edu
On Tue, 24 Jun 2003, Dirk Eddelbuettel wrote:
> On Tue, Jun 24, 2003 at 04:04:40PM -0400, Andrew Perrin wrote:
> > ii r-base 1.7.0-0.cran.1 GNU R statistical computing language and
> > apparently there's not a debian package available?
>
>
ib/R/site-library/e1071/libs/e1071.so
-rwxr-xr-x1 root staff 109247 Jun 24 15:41
/usr/local/lib/R/site-library/e1071/libs/e1071.so
> best,
> David.
>
Thanks,
Andy
>
> On 2003.06.24 21:43, Andrew Perrin wrote:
> > After upgrading to 1.7.0 under debian linux, I c
After upgrading to 1.7.0 under debian linux, I can't get e1071 working
properly.
The first problem I had was that g++-3.0 was the standard compiler but
wasn't installed, so I installed it. e1071 then installed correctly, but I
get the following:
[EMAIL PROTECTED]:~/afshome/papers/authoritarian/R$
Thanks to all who answered... this is what I was looking for:
df$rwa.sum <- df$first.agg+df$first.sub+df$first.agg
thanks to the several people who responded.
ap
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Is there a shortcut way to add a calculated variable to a data frame? For
example, I have a data frame with variables first.conv, first.sub, and
first.agg. Each cell is -1, 0, or 1. I'd like to generate variables:
rwa.sum : sum(first.agg,first.sub,first.conv)
rwa.psum : total number (0-3) of vars
John Fox was kind enough to reply, but didn't recommend IMHO the best book
on regression models: his own, John Fox, _An R and S-Plus Companion to
Applied Regression_, Sage, 2002.
ap
--
Andrew J Perrin - http://www.unc.edu/~aperri
I'm grappling with a problem and would appreciate any thoughts on it.
I'm revising a paper for resubmission to a journal. For the paper, I've
coded each "turn" in a series of conversations with several binary codes.
(A turn is one package of statements made by one speaker, starting with
the beginn
lme does this (and that is from lme) sometimes -- often when the model
> does not fit well.
>
> On Fri, 30 May 2003, Andrew Perrin wrote:
>
> > Can anyone shed any light on this?
> >
> > > doubt.demographic.pql<-glmmPQL(random = ~ 1 | groupid/participantid,
>
Can anyone shed any light on this?
> doubt.demographic.pql<-glmmPQL(random = ~ 1 | groupid/participantid,
+fixed = r.info.doubt ~
+realage + minority + female + education + income + scenario,
+data = fgdata.df[coded.resour
I think this depends on what you mean by "trend." What I would mean is
"effect of successive trials that is very unlikely to be spurious," which
is a good lay definition of statistical significance.
Given that these are multiple trials on the same subjects over time, it
seems like a mixed-effects
Yes, I do. Here's a code snippet that worked for me:
library(RPgSQL)
db.connect(dbname='fgdata')
fgdata.df<-db.read.table('mlm_data')
This assumes you have the RPgSQL library installed, and that you have
postgres running on the local machine. There's more documentation in the
RPgSQL package.
B
37 matches
Mail list logo