Thomas Lumley <[EMAIL PROTECTED]> writes:
> I think this is a bug, since it doesn't happen if factor1 isn't a factor,
> and leaving aside any question about Type III SS it seems to make it
> impossible to fit the model
>lm(vardep~factor2+factor1:factor2)
> While this model isn't terribly often
> Hello:
>
> Is there a function that will do exact logistic regressions along the
> line of LogXact?
>
No, as far as I know, there is no package to perform exact logistic
regression (namely calculate the complete permutational distribution of the
statistics of interest conditional on the ancillar
Hallo
Could anybody please help.
I have a simple linear regression model with 5 predictors. I want to
use "bootstrap residuals" to make inferences regarding beta(2)hat. After
fitting the model y=b0+b1+b2+b3+b4+b5 I tried the following:
mod <- lm(y ~ x1+x2+x3+x4+x5)
res <- resid(mod)
pred <- predic
A common approach for getting data into R from Excel is to
save the spreadsheet as a CSV file and then read it into R
using read.csv. CRAN contains references to other means of
directly linking Excel and R. You may find the CSV approach
good enough.
Regards,
Andrew C. Ward
CAPE Centre
Departme
Hello,
I am trying to use 'R' for K-means simulatio, could you please advise me how I can
read my data into a two dimesional array? Or is there any method which directly reads
the excell file? Please let me know asap.
Regards
Skanda Kallur
Cogito, Ergo Sum! Rene Descartes
___
Hello:
Is there a function that will do exact logistic regressions along the
line of LogXact?
Thanks,
ANDREW
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Greetings,
There are several editors that can call command prompts or have conficuration tools.
While learning these, I have noticed that I do not know how to execute an R program
from the command line external to the R interactive environment. It is very clearly
stated for Unix type systems,
Greetings,
I know that most people on here are strong advocates of Unix or a similar OS, but I am
interested in calling C++ or using C++ to call R. What is the preferred or the
compiler that works the best. In the documentation it appears that VC++ is
recommended, but in some posts I have see
Dear Thomas et al.,
At 05:33 PM 3/6/2003 -0800, Thomas Lumley wrote:
On Fri, 7 Mar 2003, Josef Frank wrote:
> Hello,
>
> as far as I see, R reports type I sums of squares. I'd like to get R to
> print out type III sums of squares.
>
> e.g. I have the following model:
> vardep~factor1*factor2
>
>
On Fri, 7 Mar 2003, Josef Frank wrote:
> Hello,
>
> as far as I see, R reports type I sums of squares. I'd like to get R to
> print out type III sums of squares.
>
> e.g. I have the following model:
> vardep~factor1*factor2
>
> to get the type III sum of squares for factor1 I've tried
> anova(lm(v
> From: Rolf Turner [mailto:[EMAIL PROTECTED]
>
> Andy Liaw wrote:
>
> > The long(er) answer: think harder about what question(s)
> you want answered
> > (i.e., what hypotheses you really want to test, and test
> only those). The
> > model hierarchy says that a model should not have an
> i
Hi peoples,
I'm trying to work out a function which will allow me to relace column names
on the basis of substrings within the existing names. e.g.
I'd like:
blah.Na blah2.Na blah3.Mg blah4.Mg blah5.K blah6.K
R1 x x x xxx
R2 x x
Andy Liaw wrote:
> The long(er) answer: think harder about what question(s) you want answered
> (i.e., what hypotheses you really want to test, and test only those). The
> model hierarchy says that a model should not have an interaction term
> involving a factor whose main effect is not pres
The short answer: use drop1().
The long(er) answer: think harder about what question(s) you want answered
(i.e., what hypotheses you really want to test, and test only those). The
model hierarchy says that a model should not have an interaction term
involving a factor whose main effect is not pre
Hello,
I am trying to compile the R package on a sun.
I get the error message
rbitmap.c: In function 'my_png_error':
rbitmap.c:73: structure has no member named 'jmpbuf'
rbitmap.c: In function 'R_SaveAsPng':
rbitmap.c:122: structure has no member named 'jmpbuf'
make[4]:*** [rbitmap.lo] Error 1
H
Hello,
as far as I see, R reports type I sums of squares. I'd like to get R to
print out type III sums of squares.
e.g. I have the following model:
vardep~factor1*factor2
to get the type III sum of squares for factor1 I've tried
anova(lm(vardep~factor2+factor1:factor2),lm(vardep~factor1*factor2
I've encountered some of the same issues with jaggy graphics. Here
is what I have found to work pretty well:
Short and sweet:
Use the bitmap() function with res=200 to create png files.
Some explaination:
As far as I can tell, png(), jpg(), and friends do not allow you to
set the resolution.
You may try something like this:
postscript(file="fig.ps",height=4,width=4)
layout(matrix(c(0,1,2,0,3,4),2,3,byrow=TRUE),c(0,1,1),c(1,1))
par(mar=c(5,5,2,2)+.1,mex=.6)
The default spacings are different for 2x3 and 2x2. The layout
facility allows one to cheat it out.
Good luck.
Chong Gu
> K
On Fri, 7 Mar 2003, Robin Hankin wrote:
> Kia Ora everybody.
>
> There must be an obvious answer to this, but I can't see it
>
> I want four square plots in one postscript file. The canonical answer
> would be:
>
> postscript(file="~/f.ps",width=5,height=5)
> par(pty="s",mfrow=c(2,2))
> pl
Kia Ora everybody.
There must be an obvious answer to this, but I can't see it
I want four square plots in one postscript file. The canonical answer
would be:
postscript(file="~/f.ps",width=5,height=5)
par(pty="s",mfrow=c(2,2))
plot(1:19,xlab="")
plot(1:19,xlab="")
plot(1:19,xlab="")
plot(1
>From Martin Maechler:
PS: Please guys, stop doing the following:
1. You want to post something to R-help.
2. You are too lazy to type [EMAIL PROTECTED] [the
shortest of several possible addresses]
3. you reply to another __unrelated__ posting to R-help instead.
==> all thr
> Announcing the initial release of RSessionDA.
>
> RSessionDA provides objects for interacting with R from Zope
> , a full-featured web application development system. These
> objects permit evaluation of functions in the R language using information
> in Zope. R data objects, graphics files,
[EMAIL PROTECTED] writes:
> I did not know about stepsize being an issue. I had thought that
> problems with convergence in this case were due to bad approximations of
> the finite difference gradient. I guessed that around the optimum,
> numerical errors would come to dominate the gradient calc
The simplest way is to define the method correctly, as print.myclass(x,
myarg, ...). That message is about the code, not the documentation, but
you will get a different one if the code and documentation disagree.
All print methods must include `...', and there is a section in `Writing R
Extension
On Thu, 6 Mar 2003, Dirk Eddelbuettel wrote:
> On Thu, Mar 06, 2003 at 11:53:54AM -0800, Michael Na Li wrote:
> > On Mon, 3 Mar 2003, [EMAIL PROTECTED] outgrape:
> >
> > > Rterm is a Windows application. It works fine in Windows tcsh and in
> > > Cygwin bash on Windows XP (and last time I loo
Simon Gatehouse writes:
" Like many, I fiddle while thinking. Part of my fiddling has been to
rapidly
resize the R console window back and forth by dragging with the mouse on
the
bottom right hand corner. I resize the window by a small amount rapidly
and
continually . After about 5 seconds of su
Frequently we have a print method, say print.myclass, that has a variety of arguments.
If in the .Rd file I say
\usage{
\method{print}{myclass}(x, myarg)
}
I get a warning when running R CMD chk:
* checking generic/method consistency ... WARNING
print:
function(x, ...)
print.myclass:
funct
On Thu, Mar 06, 2003 at 11:53:54AM -0800, Michael Na Li wrote:
> On Mon, 3 Mar 2003, [EMAIL PROTECTED] outgrape:
>
> > Rterm is a Windows application. It works fine in Windows tcsh and in
> > Cygwin bash on Windows XP (and last time I looked, Windows 98 too).
>
> Rterm doesn't work (interacti
This is misinformation: the font name in the PDF specifications is
ZapfDingbats, and that is what the R driver uses, as in
<<
/Type /Font
/Subtype /Type1
/Name /F6
/BaseFont /ZapfDingbats
>>
I hope you have sent a suitable bug report to the supplier of your tools.x
On Thu, 6 Mar 2003, Martin Ren
On Mon, 3 Mar 2003, [EMAIL PROTECTED] outgrape:
> Rterm is a Windows application. It works fine in Windows tcsh and in
> Cygwin bash on Windows XP (and last time I looked, Windows 98 too).
Rterm doesn't work (interactively) in Cygwin bash on our machine (Windows 2000
SP 3, R 1.6.1). (Not tha
I did not know about stepsize being an issue. I had thought that
problems with convergence in this case were due to bad approximations of
the finite difference gradient. I guessed that around the optimum,
numerical errors would come to dominate the gradient calculations,
causing convergence to f
Very often when I print a data frame (particularly when sink()-ing to
a file I do NOT want the row labels (which are in such cases usually
1, 2, ... nrow(x), where x is the data frame in question). I can of
course edit these out ``by hand'', but that's a bit of a pain in the
pohutukawa.
I recentl
Daniel A. Powers wrote:
R-list --
I am working with some SPlus code and get the following error
Error in terms.default(formula, data = data) :
no terms component
Can someone point me to a fix? The pasted formula looks ok
when printed during the procedure.
Without a more detailed descripti
R and S-PLUS handle terms very differently.
If this were a formula, terms would dispatch to term.formula, so
presumably it is not. Do you need an as.formula() call?
On Thu, 6 Mar 2003, Daniel A. Powers wrote:
> I am working with some SPlus code and get the following error
>
> Error in terms.de
I had the same problem and came to this solution:
- use pdf() to create a pdf file
- open PDF in Freehand,
during open replace fonts: ZapfDingsbat for Zapf Dingsbat
- save as PDF or EPS and import this into word - bingo
The pdf files created by pdf() seem less than perfect - on my system
(
R-list --
I am working with some SPlus code and get the following error
Error in terms.default(formula, data = data) :
no terms component
Can someone point me to a fix? The pasted formula looks ok
when printed during the procedure.
Thanks,
Dan
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
[EMAIL PROTECTED] (Michael A. Miller) writes:
> Dear R Gang,
>
> I'm interested in using R and the nls package for fitting kinetic
> models. I'm having some difficulty getting a model specified for
> nls though. The math for the model that I want to fit is
>
>dg(t)/dt = K1 f(t) - k2 g(t)
>
I've found Word imports png pretty well. David Brahm recently suggested
playing with the resolution and so on -- you could look it up in the mail
list archives -- but I've gotten acceptable results (on simple graphs) with
the defaults.
Hope this helps,
Matt Wiener
-Original Message-
From
On Thu, Mar 06, 2003 at 10:08:34AM -0800, Eric Fruits wrote:
[four permutations on pdf() snipped]
Have you tried other graphics drivers? help(png) and help(jpeg) might
be other options. I'd suggest playing with the "height" and "width"
options.
It sounds like Word for OS-X makes whatever you'r
"janet rosenbaum" <[EMAIL PROTECTED]> writes:
> 2. I also have a tcltk in R question.
>
> How do you use the command tkoptionmenu in R?
>
> The following are both valid syntax and both create the desired menu,
> but neither changes the variable fsep.
>
> septype.menu <- tkoptionmenu(septy
Dear R Gang,
I'm interested in using R and the nls package for fitting kinetic
models. I'm having some difficulty getting a model specified for
nls though. The math for the model that I want to fit is
dg(t)/dt = K1 f(t) - k2 g(t)
where g(t) and f(t) are measured data at a sequence of times
After trying numerous options, I'm just about at my wits end.
The most frequent suggestion was to export to a postscript or PDF file
and import that into Word.
However, no matter what I did or how I did it, the results were
extraordinarily ugly and somewhat time-consuming.
What I've tried so f
Regarding "1", both "cat(x)" and "print(x)" have worked for me,
depending on the class of "x". To get extra labeling, sometimes I have
used
print(list(x=x))
and variants on that.
Hope this helps.
Spencer Graves
janet rosenbaum wrote:
I have two questions: one so easy that I apologise for it
I have two questions: one so easy that I apologise for it in advance
and one maybe more obscure:
1. Interactively, typing the variable name will cause its value to print.
How do I print out a variable's value inside of a script?
I've tried cat("Field separator" fsep) and variations thereon, an
Christoph Lehmann <[EMAIL PROTECTED]> writes:
> is there any package or method which enables to compute a linear
> regression with the leas absolute value fit-criterion?
Try rq() from the quantreg package.
--
O__ Peter Dalgaard Blegdamsvej 3
c/ /'_ --- Dept. of Biostati
is there any package or method which enables to compute a linear
regression with the leas absolute value fit-criterion?
thanks
christoph
--
Christoph LehmannPhone: ++41 31 930 93 83
Department of Psychiatric NeurophysiologyMobile: ++41 76 570 28 00
University H
Hi all,
I have problems in a dataframe variables types.
Look:
from a loop function:
for(...){
...
dados.fin <- rbind(dados.fin, c(L=j, A=j^2,
Nsp=nsps,
N=length(amosfin$SP),
AmT="am",NAm=nam,
Would tkdestroy() work for you?
>Date: Thu, 6 Mar 2003 11:33:05 -0300 (ART)
>From: <[EMAIL PROTECTED]>
>To: [EMAIL PROTECTED]
>MIME-Version: 1.0
>Content-Transfer-Encoding: 8bit
>X-Virus-Scanned: by amavisd-milter (http://amavis.org/)
>X-Virus-Scanned: by amavisd-milter (http://amavis.org/)
>X-Vir
Sorry, my question is if exists a command tkexit in R, i need a command how
exit of tcl for close an aplication. How to use him. Thank
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Hello, i want post my mail in r-announce. [EMAIL PROTECTED] thanks
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Hello, I want post my mail [EMAIL PROTECTED]
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[EMAIL PROTECTED] wrote:
> zdump manipulates its environment directly but is otherwise the same code.
> Solaris and glibc define putenv slightly differently. There are
> two lines like
>
> char buff[20];
>
> in src/main/datetime.c, and if you change those to
>
> static char buff[2
zdump manipulates its environment directly but is otherwise the same code.
Solaris and glibc define putenv slightly differently. There are
two lines like
char buff[20];
in src/main/datetime.c, and if you change those to
static char buff[200];
it should work (increasing the len
On 6 Mar 2003 at 10:38, rob foxall (IFR) wrote:
You can use linear.hypothesis() from the
package car (on CRAN).
Kjetil Halvorsen
> Hello all,
>
> A really noddy question for you all: I'm trying without success to do
> some subhypothesis testing. Using simple anova model, with a toy
Shingo Ichii <[EMAIL PROTECTED]> writes:
> Hi,
>
> > I just changed to FreeBSD platform, and want to install R on it. I use
> > FreeBSD 5.0 and install nearly all packages on the machine. When I use
> > ports to install R, (cd /usr/ports/math/R-letter, and then type make) I
> > got the following
Oops, sorry folks, I didn't mean to send any of these emails to the R
list, I thought I was having a private discussion...I'm going to have
to see what I did wrong.
Dave
iOn Thu, Mar 06, 2003 at 11:35:35AM +0100, Peter Dalgaard BSA wrote:
> [EMAIL PROTECTED] writes:
>
> > Ah, but the interes
"rob foxall (IFR)" <[EMAIL PROTECTED]> writes:
> Hello all,
>
> A really noddy question for you all: I'm trying without success to do
> some subhypothesis testing. Using simple anova model, with a toy dataset from a
> book. I have four factors A,B,C,D, and wish to test mu_C = mu_D.
Hi,
> I just changed to FreeBSD platform, and want to install R on it. I use
> FreeBSD 5.0 and install nearly all packages on the machine. When I use
> ports to install R, (cd /usr/ports/math/R-letter, and then type make) I
> got the following error information.
> ../../../../library/methods/libs/
Hello all,
A really noddy question for you all: I'm trying without success to do some
subhypothesis testing. Using simple anova model, with a toy dataset from a book. I
have four factors A,B,C,D, and wish to test mu_C = mu_D. This is what I have tried:
> contrasts(infants$group,h
[EMAIL PROTECTED] writes:
> Ah, but the interesting thing is that they are coming out with a 'clam'
> version like the 5MX. Details are limited at the moment, but that could
> mean the combination of 5MX usability with a supported linux distro. I
> am drooling in anticipation. Sounds like a "I'
Ah, but the interesting thing is that they are coming out with a 'clam'
version like the 5MX. Details are limited at the moment, but that could
mean the combination of 5MX usability with a supported linux distro. I
am drooling in anticipation. Sounds like a "I've finally finished my
PhD and dese
Hi,
I need to test multiple correlation coefficient in a same time,
what's the best test for that and are should I do under R ?
Thanks
--
Cordialement
Emmanuel POIZOT
Cnam/Intechmer
Digue de Collignon
50110 Tourlaville
Tél : (33)(0)2 33 88 73 42
Fax : (33)(
Or even:
which(lv)[1]
as lv is already a logical vector
Giles
On 2003.03.05 22:44 Spencer Graves wrote:
Also:
which(lv == TRUE)[1]
Spencer Graves
Marc Schwartz wrote:
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jason Liao
Sent: Wednesday, Marc
> "Richard" == Richard A Bilonick <[EMAIL PROTECTED]>
> on Wed, 05 Mar 2003 12:01:01 -0500 writes:
Richard> I think I've figured out the use of "filter" in the ts package, at least
Richard> for a simple AR model. I simulated a simple AR time series, modeled it
Richard> us
It's a Linux issue. Solaris gives:
> Sys.time()
[1] "2003-03-06 08:27:23 GMT"
> as.POSIXlt(Sys.time(), "HST")
[1] "2003-03-05 22:27:40 HST"
which looks right to me. Past experience suggests that Solaris's POSIX
conformance is much better than glibc's. Since my RH7.2 box does it too,
I will loo
table() turns its argument into a factor, as it is documented to work on
factors, only, and
> factor(x$date)
[1] 1034809200 1034809200 1034809200 1034809200 944611200 944611200
Levels: 944611200 1034809200
That's because unique.default does not know about POSIXct objects (nor
indeed many othe
66 matches
Mail list logo