Dear all,
I wonder if there are some function or package in R which can do the
iterative proportional fitting.
In the exponential model
f(y1,...yn)=exp(a'yi+b'(yi*yj)+.+c'(y1*...*yn)+constance),
instead of the canonical parameters, I use maginal probability instead of a
and log odds ratio
names() is only defined for vectors and lists and factors are
neither. See ?vector and ?names for more info.
---
From: djw1005 at cam.ac.uk
Subject: [R] Factor names & levels
When I alter the levels of a factor, why does it alter the names too?
f <- factor(c(A="one",B="two",C="one",D="one"
One of R's goals has always been to run on minimal hardware, and we say
for example (src/gnuwin32/CHANGES for rw1070)
This version of R needs more memory and is slower to start up, because
it loads more packages by default. This is only likely to be a
concern on machines with 16Mb of memor
Based on an off list email conversation, I had I am concerned that
my original email was not sufficiently clear.
Recall that I wanted to use a for loop to iterate over the rows of
a dataframe without using indices. Its easy to do this over
the columns (for(v in df) ...) but not for rows.
Wha
First, xreg is documented as a vector or matrix, not a data frame.
Second, there may be a scoping problem, so try a more informative name
than `y' when you do the fit.
Third, R has debugging facilities, so please use them to find out
precisely what is wrong.
On Thu, 18 Dec 2003, Yves wrote:
>
> li <- list(1, 2, 3)
> li
[[1]]
[1] 1
[[2]]
[1] 2
[[3]]
[1] 3
> paste(unlist(li), collapse=" ")
[1] "1 2 3"
> length( paste(unlist(li), collapse=" ") )
[1] 1
--
Adaikalavan Ramasamy
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Christopher Brown
To all
I have some simple questions pertaining to weights used in regression.
If the variability of the dependent variable (y) is a function of the magnitude of
predictor
variable (x), can the use of weights give an appropriate answer to the regression
parameters
and the std errors?
Assume th
On Thu, Dec 18, 2003 at 05:37:49PM -0500, David James wrote:
> AFAIK the only way to connect to Sybase from R is with the RODBC
> package, so you need to have an ODBC driver for Sybase -- either
> one provided by Sybase or the FreeTDS ODBC driver from www.freetds.org.
Seconded -- we did that recen
You could just convert your binary spatial data to numeric 0/1 or -1/1 and
give it to Mclust. That would violate the assumptions of the Gaussian
model in Mclust, so you should be very careful about interpreting the
results. However, if the results are at least as "interesting" as those
you ge
One can perform a for loop without indices over the columns
of a dataframe like this:
for( v in df ) ... some statements involving v ...
Is there some way to do this for rows other than using indices:
for( i in 1:nrow(df) ) ... some statements involving df[i,] ...
If the dataframe had o
On Fri, 19 Dec 2003, Murray Jorgensen wrote:
> Thomas W Blackwell wrote:
>
> > [...]
> > Why not simply use dist() and hclust() ? Starting with
> > presence/absence data, what could mclust() possibly do that
> > is different from hclust() ?
>
> Um, fit a statistical model.
Yes, but ellips
Thomas W Blackwell wrote:
[...]
Why not simply use dist() and hclust() ? Starting with
presence/absence data, what could mclust() possibly do that
is different from hclust() ?
Um, fit a statistical model.
- tom blackwell - u michigan medical school - ann arbor -
_
On Thu, 18 Dec 2003, Jarrod Hadfield wrote:
> Dear All,
>
> I have spatial data (presence/absence for 4000 squares) on 250 bird
> species and would like to use a model-based clustering technique to
> test for species associations. Is there any way of passing a
> distance/correlation matrix to mcl
Frank -
I recall some recent discussion on this list about installing
Design on a Mac. Try the mail archive for a message from
Frank Harrell, the package author and maintainer within the
last five weeks. Your question sounds very similar to the
question I remember . . . but mine is only a human
AFAIK the only way to connect to Sybase from R is with the RODBC
package, so you need to have an ODBC driver for Sybase -- either
one provided by Sybase or the FreeTDS ODBC driver from www.freetds.org.
Regards,
--
David
Peter McMahan wrote:
> Hi,
> I'm trying to get R connected to Sybase on a Li
Check out
?summary.formula
in package Hmisc.
---
Date: Thu, 18 Dec 2003 13:27:39 -0600
From: Paul Johnson <[EMAIL PROTECTED]>
To: '[EMAIL PROTECTED]' <[EMAIL PROTECTED]>
Subject: [R] diagnostic information in glm. How about N of missing observations?
I handed out some results from g
Hi,
I'm trying to get R connected to Sybase on a Linux
machine. I'm relatively new to databases,
particularly the connectivity side.
Do I have to use RODBC and install ODBC support on the
system (a huge pain with Sybase, as you have to
install a scaled-down version of Sybase itself on the
system t
Whoops. That should be a "&&" where I put a "||".
naPrint <- function(model,...) {
if(!inherits(model,"lm") && !inherits(model,"glm"))
stop("no support for class ",class(model),"\n")
if(!is.null(model$na.action)) {
action <- paste("na.action:",attr(model$na.action,"class")
Rolf Turner wrote:
Paul E. Johnson wrote:
I handed out some results from glm() and the students ask "how many
observations were dropped due to missing values"?
How would I know?
...
It
would be ***VERY*** easy to write a wrapper for glm() to include
the number of dropped observations if you
Hi all there
I am enjoying R since 2 weeks and I come to my first deadlock, il am trying
to use predict.Arima in the ts package.
I get a "Error in cbind(...) : cannot create a matrix from these types"
-- Start R session -
> fitdiv <- arima(dat
Hi all there
I am enjoying R since 2 weeks and I come to my first deadlock, il am trying
to use predict.Arima in the ts package.
I get a "Error in cbind(...) : cannot create a matrix from these types"
-- Start R session -
> fitdiv <- arima(data
Dear All,
I have spatial data (presence/absence for 4000 squares) on 250 bird
species and would like to use a model-based clustering technique to
test for species associations. Is there any way of passing a
distance/correlation matrix to mclust as with hclust, rather than the
actual data? Or
Dear Angel,
> I needed a symbolic and algebraic computation software. The best solution
> I've found so far is maxima computer algebra system.
Maybe there is also the CAS MuPAD of interest, which has the look and feel of
maple, works under Linux, Mac and Windows - and has a free academic version
I'm wondering if someone has installed the Design library on OS-X.
I have the Developer tools installed (not the OS X SDK kit) but after
downloading the library via RAqua package manager I get the following
errors:
Installation of package Design had a non-zero exit status in :
install.packages(u
There should be an element of the fitted model object called `na.action'
which (if you used the default na.omit()) should tell you which
observations were dropped.
-roger
Paul Johnson wrote:
I handed out some results from glm() and the students ask "how many
observations were dropped due to mi
>Hi R-Masters
I found a strange error in fitdistr():
In case of VR Version 7.1-11
k21stsList<-c(0.76697,0.57642,0.75938,0.82616,0.93706,0.77377,0.58923,0.37157,0.60796,1.00070,0.97529,0.62858,0.63504,0.68697,0.61714,0.75227,1.16390,0.66702,0.83578)
fitdistr(k21stsList, "normal",list(mean = 0.
Paul E. Johnson wrote:
> I handed out some results from glm() and the students ask "how many
> observations were dropped due to missing values"?
>
> How would I know?
>
> In other stat programs, the results will typically include N and
> the number dropped because of missings. Without going
I handed out some results from glm() and the students ask "how many
observations were dropped due to missing values"?
How would I know?
In other stat programs, the results will typically include N and the
number dropped because of missings. Without going back to R and
fiddling about to find
Wednesday saw the release of both the movie "Lord of the Rings: Return
of the King" (LotR:RotK) and the Linux 2.6.0 kernel. Those who enjoy
geek humor should read the discussion on slashdot.org from people who are
lost in the agonies of trying to decide whether to go see the movie or
to compile th
What is intended to happen is for it silently to use font 1. I am
committing a fix to ensure that happens in the one spot that got missed.
On Thu, 18 Dec 2003, Gavin Simpson wrote:
> Dear Andy, Duncan and Brian
>
> Cheers for your replies
>
> That font.main thing is something I didn't pick up
Dear Andy, Duncan and Brian
Cheers for your replies
That font.main thing is something I didn't pick up on when I cropped my
colleagues code down to a small example.
font.main works fine with pointsize = 8, *if* you set font.main to one
of the values given in the entry for font in ?par. The cod
What do you think font.main=16 does? postscript() only has 5 fonts.
Fortunately your example also crashes on Linux so was fairly easy to track
down.
The windows devices do have more fonts (19 by default).
On Thu, 18 Dec 2003, Gavin Simpson wrote:
> Dear List,
>
> My colleague has been having
On Thu, 18 Dec 2003 18:02:07 +, Gavin Simpson
<[EMAIL PROTECTED]> wrote :
>Dear List,
>
>My colleague has been having a problem with the following data and
>plotting commands. The example below is part of a larger set of plots,
>but I've isolated the problem to this example using this small
The font.main = 16 in par() is having problems cooperating with
pointsize = 8 in postscript().
One of them has to go and then the code is happy. This is probably a
feature.
HTH, Andy
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Gavin Simpson
Sent: Th
Dear List,
My colleague has been having a problem with the following data and
plotting commands. The example below is part of a larger set of plots,
but I've isolated the problem to this example using this small dataset
(below), which kills rgui consistently. My version info
> version
This is probably an easy question.
I want to join the elements of a list/array into a single scalar value. How can I do
this?
More Background:
I have sql queries stored in external flat text files. When I read these queries into
R using the normal functions, I get a list where each line is
Brian,
Thanks so much for your comments. Like I said, I am pretty much a novice at this
idea. My statistics degree emphasized a lot of theoretical knowledge, and the idea of
simulation and resampling was never taught. So, now that I am more in an applied
field, I realize the need to educate
The key is the `keep.order' argument to terms.formula:
> options(contrasts=c("contr.helmert", "contr.poly"))
> library(MASS)
> aov(yield ~ block + N * P + K, npk)
Call:
aov(formula = yield ~ block + N * P + K, data = npk)
Terms:
blockNPK N:P Resi
I had made a second subsequent suggestion to address this. See:
https://www.stat.math.ethz.ch/pipermail/r-help/2003-December/042234.html
---
Date: Thu, 18 Dec 2003 07:48:26 -0800 (PST)
From: Thomas Lumley <[EMAIL PROTECTED]>
To: Gabor Grothendieck <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED
On Wed, 17 Dec 2003, Gabor Grothendieck wrote:
>
>
> In rereading this one idea occurred to me. What if the entire R help
> system were turned into a wiki? That is,
>
> ?whatever
>
> would take you to the help page, but not on your computer --
> rather to the same page on the wiki.
This seems
On Thu, 18 Dec 2003, juli g. pausas wrote:
> Dear colleges,
> I do not understand the following behaviour:
>
> > aa <- data.frame(a1= 1:10, a2= c(rep(NA, 5), 1:5) )
> > aa[!aa$a2==1, ] # removing rows with a2==1
> a1 a2
> NA NA NA
> NA.1 NA NA
> NA.2 NA NA
> NA.3 NA NA
> NA.4 NA NA
> 7
Alternatively, C-c TAB might work.
"Pascal A. Niklaus" <[EMAIL PROTECTED]> writes:
> Sean Davis wrote:
>
>>I don't think that this works within the R process buffer, but I could be
>>wrong. Does the documentation say that it should somewhere?
>>
>>Sean
>>
> The ESS doc says (section 3.2):
>
>
On Thu, 18 Dec 2003, Mulholland, Tom wrote:
>
> 1 An assumption on my part is that there is fundamental agreement that
> the document is the best source for advice on how to ask questions of
> this list
I don't think its the best possible source. It may well be the best
existing source.
On bug re
If DF = a data.frame with 65000 rows and k columns, the following
will do what I read in your question:
lapply(DF, table)
See, e.g., Venables and Ripley, Modern Applied Statistics with S
(Springer, pp. 33-34 in the 4th edition, 2002).
hope this helps.
Perez Martin, Agus
On Wed, 17 Dec 2003, Tony Plate wrote:
> Another way to approach this is to first massage the data into a more
> regular format. This may or may not be simpler or faster than other
> solutions suggested.
You could also use the reshape() command to do the massaging
-thomas
> > x <- rea
There is also the open source program Pari:
Quote:
I would like to announce the release of pari-2.2.7.ALPHA. The sources can be
obtained using the old address
ftp://megrez.math.u-bordeaux.fr/pub/pari/pari-alpha.tgz
or (preferably) through the new website:
http://pari.math.u-bordeaux.fr/dow
Is there a way to prevent the re-ordering of factors by aov? I do have a
three-way interaction that I do want to fit before a two-way interaction
(different factors, so they are not nested), but R moves the two-way
interaction to the front. I know it generally makes sense to fit the
two-way int
Sean Davis wrote:
I don't think that this works within the R process buffer, but I could be
wrong. Does the documentation say that it should somewhere?
Sean
The ESS doc says (section 3.2):
Completion is also provided over file names, which is particularly
useful when using S functions such as
Hi Allan,
I think the function lmp() of the package normalp could be useful to you.
Regards,
Elio
allan clark wrote:
Hi all
Just wondering whether one can undertake Lp norm estimation (a type of
regression analysis) in R?
i.e.
argmin S ( | y(i) - x(i)b | ^p )
where:
* S is the summation over
Allan -
Brian Ripley's implementation of one of the more useful Lp norms is:
library("lqs")
help("lqs")
This is very highly recommended for practical data analysis.
- tom blackwell - u michigan medical school - ann arbor -
On Thu, 18 Dec 2003, allan clark wrote:
> Hi all
>
> Just wonde
Roger Koenker wrote:
> One should doubt everything, even logic, and in doubt lies possibility.
If we doubt logic, discourse becomes impossible.
cheers,
Rolf Turner
__
[EMAIL PRO
On Thu, 18 Dec 2003 15:03:04 +0100 juli g. pausas wrote:
> Dear colleges,
> I do not understand the following behaviour:
>
> > aa <- data.frame(a1= 1:10, a2= c(rep(NA, 5), 1:5) )
> > aa
>a1 a2
> 1 1 NA
> 2 2 NA
> 3 3 NA
> 4 4 NA
> 5 5 NA
> 6 6 1
> 7 7 2
> 8 8 3
> 9 9 4
>
Take a look at what (aa$a2 == 1) returns and it may clear things up.
Try
aa[-which(aa$a2 == 1), ]
or
subset(aa, a2 != 1 | is.na(a2))
HTH,
Sundar
juli g. pausas wrote:
Dear colleges,
I do not understand the following behaviour:
aa <- data.frame(a1= 1:10, a2= c(rep(NA, 5), 1:5) )
aa
a1 a2
1
On Thu, Dec 18, 2003 at 12:04:31PM +0100, Peter Dalgaard wrote:
> Göran Broström <[EMAIL PROTECTED]> writes:
>
> > Dear R-helpers,
> >
> > In a data set I got from a medical doctor there are six treatment groups
> > and (about) 5 bivariate responses in each group. Using 'manova', it is
> > easy t
Hi all,
Using Xemacs/ESS (Xemacs 21.4-12/ESS 5.1.24), I noticed that tab
completion does not work for file names.
df <- read.csv("xx [TAB]
doesn't complete the file name.
After hitting TAB, I get "last thing matched was not a buffer" in the
status bar. Any idea what may go wrong?
Thanks
One should doubt everything, even logic, and in doubt lies possibility.
url:www.econ.uiuc.edu/~roger/my.htmlRoger Koenker
email [EMAIL PROTECTED] Department of Economics
vox:217-333-4558University of Illinois
fax:217-244-6678
Dear colleges,
I do not understand the following behaviour:
aa <- data.frame(a1= 1:10, a2= c(rep(NA, 5), 1:5) )
aa
a1 a2
1 1 NA
2 2 NA
3 3 NA
4 4 NA
5 5 NA
6 6 1
7 7 2
8 8 3
9 9 4
10 10 5
aa[!aa$a2==1, ] # removing rows with a2==1
a1 a2
NA NA NA
NA.1 NA NA
NA.2 NA NA
Hello R-List
I use DClusters package (I work in a cancer regestry). I have 2 questions
about it:
1-how is it possible to get back the bootstrap pValue? I mean the pValue of
the calculated statistic with respect of the distribution of this statistic
under the null hypothesis.
2-how is it possible
Roger Koenker wrote:
Everything is possible, some things are just more difficult than others.
Is that so? In which case, how can I construct a square with the same
area as a given circle using only the compass() and ruler() functions?
:)
Baz
PS redefining compass() and ruler() not allowed...
If your dataset contains integer or limited possible unique numbers only I find the
following more concise.
m <- matrix( rpois(60, 5), nc=6 )
apply( m , 2, function(x) table( factor(x, levels=0:max(m))) )
If your dataset has continous or lots of unique numbers you may wish to consider only
th
> From: Prof Brian Ripley
[snip]
> As a matter of terminology, this is not resampling as usually
> defined, so
> I do wonder exactly what it is you are after. For resampling
> in the usual
> sense, I would echo Jason's recommendation of Davison and
> Hinkley's CUP book.
Or perhaps at a gen
On Thu, 18 Dec 2003 11:19:20 +0100, "Angel" <[EMAIL PROTECTED]>
wrote :
>I needed a symbolic and algebraic computation software. The best solution
>I've found so far is maxima computer algebra system.
>I was wondering if there is any R package that has similar features as
>maxima's or somebody cou
Everything is possible, some things are just more difficult than others. It is a
little
tricky to find the point of intersection of the voronoi edges with the frame of the
diagram. Your basic request is quite straightforward though. Here is a function
that plots, by looping through the sites an
A 14:57 17/12/2003 +0100, Martin Wegmann a écrit:
Hello,
Roger Bivand wrote:
> appropriate light. One basic characteristic seems to be that if the
> question does indicate seriousness about trying to analyse data, respect
> for the task at hand, then predictably lots of good advice comes quickly.
Dear List-Members,
my name is Klemens Barfus and I am quite new to this list and to working
with R.
I try to generate Vornoi-Polygons with the Tripack package. I would like to
have a table with the sites xy and the surrounding nodes of the
Voronoi-Polygons. The area shall be delimited by a defined
Hi all
Just wondering whether one can undertake Lp norm estimation (a type of
regression analysis) in R?
i.e.
argmin S ( | y(i) - x(i)b | ^p )
where:
* S is the summation over observation i= 1,2,...,n
* y is a vector of n observations
* x is an n by p matrix of explanatory variables
Hi Andrea.
When we run Perl from within R, any Perl module that has associated
C code (such as modperl) needs a special piece of bootstrapping code
to be used within this embedded case. Now in your case, you might
only be calling R from within Perl, but neverthless, we put this
bootstrapping
If anyone can help it would be very much apreciated...
System RedHat 9
R installed as
# ./configure --enable-R-shlib
# make
# make install
and R seems to work fine ...
Then I do
# R INSTALL --clean --configure-args='--with-in-perl'
RSPerl_0.5-7.tar.gz
# export R_HOME=/usr/local/lib/R
# cd
Dear all,
Thanks to Thomas Lumley, Peter Wolf and Gabor Grothendieck for their reply
The problem was to to apply a bivariate function (such as cor(), for
instance) to each combination of colums of a given matrix.
Below there are four possible solutions (the original message is also posted
below).
Göran Broström <[EMAIL PROTECTED]> writes:
> Dear R-helpers,
>
> In a data set I got from a medical doctor there are six treatment groups
> and (about) 5 bivariate responses in each group. Using 'manova', it is
> easy to see significant differences in treatment effects, but the doctor
> is more i
Thank you very much!
I'm using par(new = TRUE) because in my real case, the 2 plots have
different ylim (different y-scale).
I got what I wanted by using the same xlim in the barplot and in the
plot, as suggested by Peter.
My real case:
par(mar= c(7, 4, 5, 5) + 0.1)
area <- c(136, 3426, 559
Dear R-helpers,
Does anybody know, whether there is an option to tell aov/anova, or do
something similar, to get a Huynh-Feldt correction of dfs in the aov/anova
function?
Thanks in advance!
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.eth
Dear R-helpers,
In a data set I got from a medical doctor there are six treatment groups
and (about) 5 bivariate responses in each group. Using 'manova', it is
easy to see significant differences in treatment effects, but the doctor
is more interested in the correlation between the two responses (
"juli g. pausas" <[EMAIL PROTECTED]> writes:
> Dear colleges,
> I'm trying to combine a barplot and a plot in a single figure as follows:
>
> data <- 1:6
> t <- barplot(data, axes=F)
> par(new= T)
> plot(t, data, type="b")
>
> However, as you can see in the example, the dots of the second plot d
> "juli" == juli g pausas <[EMAIL PROTECTED]>
> on Thu, 18 Dec 2003 10:54:08 +0100 writes:
juli> Dear colleges,
juli> I'm trying to combine a barplot and a plot in a single figure as follows:
juli> data <- 1:6
juli> t <- barplot(data, axes=F)
juli> par(new= T)
I needed a symbolic and algebraic computation software. The best solution
I've found so far is maxima computer algebra system.
I was wondering if there is any R package that has similar features as
maxima's or somebody could point me to reference manuals on how to use R for
this purpose.
Thanks,
An
Dear colleges,
I'm trying to combine a barplot and a plot in a single figure as follows:
data <- 1:6
t <- barplot(data, axes=F)
par(new= T)
plot(t, data, type="b")
However, as you can see in the example, the dots of the second plot do
not fall in the midpoint of the bars in the first. Any trick fo
> "Tony" == Tony Plate <[EMAIL PROTECTED]>
> on Wed, 17 Dec 2003 18:04:34 -0700 writes:
Tony> The suggestions of Tom (posting guide) and Andy (Eric
Tony> Raymond's "How To Ask Questions The Smart Way") are
Tony> both good. Perhaps a good place to put an actual
Tony> po
Have a look at 'table'
To compute for all columns of your dataset, combine with 'apply':
> data(iris)
> apply(iris,2,table)
[...]
$Petal.Width
0.1 0.2 0.3 0.4 0.5 0.6 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2
2.3 2.4 2.5
5 29 7 7 1 1 7 3 5 13 8 12 4 2 12 5
Hello UseRs:
Excuses for my english.
I have a dataset with 65000 records and I'd like to make a summary where I
can view all the values (with the number of times that it repeats) that
there are each column of my dataset.
I tried with summary( ), str( ), but nothing gives me the result that I am
lo
On Wed, 17 Dec 2003, Brandon Vaughn wrote:
> Thanks to everyone who wrote in with suggestions. I will check out the
> books mentioned.
>
> The book I mentioned "Resampling: The New Statistics" is actually available
> free online at:
>
> http://www.resample.com/content/text/index.shtml
>
> It s
81 matches
Mail list logo