These normally occur (at that point) from having written off one end of an
array. (If you are using .C/.Fortran, they try to copy back the
arguments.) Compiling with bounds checking turned on can help, at least
with Fortran.
On Wed, 12 May 2004, Ingmar Visser wrote:
I'm building a package using
Don't understand... What patch do you expect from me? And, I thought my analysis is
quite reproducable and thorough. It is im my situation, though.
I'm not patching something, or writing/issueing R-code. I'm 'just' an R-user with a
problem/question.
In short: using underscores in
Hi,
I would like to do a MannWitney test.
Can anyone help me with the propper command?
Thanks,
Margarida
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
my apologies; this messages shouldnt have been send to the list. I did reply to Brian
directly.
Sorry for your time.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Hoeven, Maarten
van der
Sent: Wednesday, May 12, 2004 3:04 PM
To: [EMAIL PROTECTED]
Hi,
did you look at wilcox.test?
Daniel
-Ursprüngliche Nachricht-
Von: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Im Auftrag von Margarida Júlia
Rodrigues Igreja
Gesendet: Mittwoch, 12. Mai 2004 15:13
An: [EMAIL PROTECTED]
Betreff: [R] mannwitney
Hi,
I would like to do a MannWitney
See ?wilcox.test. In particular, see the details section.
Margarida Júlia Rodrigues Igreja wrote:
I would like to do a MannWitney test.
Can anyone help me with the propper command?
--
Chuck Cleland, Ph.D.
NDRI, Inc.
71 West 23rd Street, 8th floor
New York, NY 10010
tel: (212) 845-4495 (Tu, Th)
DATA
[,1]
[1,] -2.0314185
[2,] -0.3594015
[3,] -1.3225832
[4,] 1.5050892
[5,] -0.4535558
[6,] 3.6335036
[7,] 5.1965750
[8,] 4.2923129
[9,] 6.0913187
[10,] 4.9131070
RESA
[,1] [,2] [,3] [,4] [,5]
[1,] -2.0314185
Is there a function in R for the computation of test power (and sample
size) when comparing two proportions from dependent samples.
Can power.prot.test be used or does it require independent samples?
Oddmund Nordgård
**
Oddmund Nordgård
Department
Thank you to David James for the solution to this one!
David James wrote:
The workaround in the README.Oracle9 files is intended for the symbol
sqlda (I think), but the sqlctx symbol apparently is defined in
-lclntsh.
So, could you try the following?
R CMD INSTALL
Sarò fuori ufficio a partire dal 12/5/04 fino al 17/5/04
Non so se avrò accesso alla posta elettronica.
I'll be away from the 12th to the 17th of May and I'm not sure I'll be able to
read my mail
[[alternative HTML version deleted]]
__
I have troubles installing the SparseM package on my Debian stable
Linux system.
Debian's version of R is:
platform i386-pc-linux-gnu
arch i386
os linux-gnu
system i386, linux-gnu
status
major1
minor5.1
year 2002
month06
day 17
language R
This is the installation
On Wed, 12 May 2004, [iso-8859-1] Göran Broström wrote:
Is it the data? Let's try 'coxreg' (eha):
---
Call:
coxreg(formula = Surv(yrs2, ratify) ~ haz.wst + pol.free, data = dat)
Covariate Mean CoefRR
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
What R functionnalities are there to do missing values imputation (substantial
proportion of missing data)?
I would prefer to use maximum likelihood methods ; is the EM algorithm implemented? in
which package?
Thanks
Anne
Anne Piotet
You need to find a version of SparseM in the CRAN archive of comparable
age to your R version (nearly two years),
*or*, much better, update your R to 1.9.0 (or the current patched version)
Hint to maintainers of packages using S4 methods: you almost certainly
need to declare a dependence on a
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Anne Piotet wrote:
What R functionnalities are there to do missing values imputation
(substantial proportion of missing data)? I would prefer to use
maximum likelihood methods ; is the EM algorithm implemented? in
which package?
The so-called ``EM algorithm'' is ***NOT*** an
Hi,
I would like to do a quisquared test, can you help me with the command?
Margarida
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Sorry, last attempt, before contacting our computer people.
Well, I am not sure what was wrong but my reply to Taemyong Choi's question
was somehow mismatched so I try again.
Hi
On 12 May 2004 at 22:35, Taemyong Choi wrote:
DATA
[,1]
[1,] -2.0314185
You need to find a version of SparseM in the CRAN archive of comparable
age to your R version (nearly two years),
*or*, much better, update your R to 1.9.0 (or the current patched
version)
What you said. :)
the current R packages available in debian unstable work VERY well.
as a bonus,
On Wed, May 12, 2004 at 10:10:56AM -0500, elijah wright wrote:
You need to find a version of SparseM in the CRAN archive of comparable
age to your R version (nearly two years),
*or*, much better, update your R to 1.9.0 (or the current patched
version)
CRAN has Debian packages of R
Hello,
I cannot find a function that forms the identity matrix from a factor, is
there any?
Patrik Waldmann###
This message has been scanned by F-Secure
Anti-Virus for Microsoft Exchange.
###
Margarida Júlia Rodrigues Igreja wrote:
Hi,
I would like to do a quisquared test, can you help me with the command?
I guess you are looking for
?chisq.test
Uwe Ligges
Margarida
__
[EMAIL PROTECTED] mailing list
Patrik Waldmann wrote:
Hello,
I cannot find a function that forms the identity matrix from a factor, is
there any?
For the identity matrix in general see ?diag.
What do you expect to get when forming a identity matrix from a factor?
Uwe Ligges
Patrik
At 14:12 2004-05-12 +0100, you wrote:
Hi,
I would like to do a MannWitney test.
Can anyone help me with the proper command?
Thanks,
Margarida
You may use wilcox.test or qwilcox, but beware of some subtleties, which
can result in a bit of confusion (see the below comments and it's
discussion in
Hello again,
I was too quick before. What I was looking for was a function that
constructs the design (or incidence) matrix (X in a linear model) from a
factor. Uwe Ligges suggested using model.matrix and this does almost what I
want, but it is first necessary to construct a data variable. It
For your first question the simplest way to handle it is to
sample the indices rather than the data. For example,
x - seq(10,200,10) # data
ix - sample(length(x),10,rep=T) # 10 samples from 1:length(x)
s - x[ix] # samples
Now ix has the indices of s.
Taemyong Choi tmchoi at
I've got a newbie question and I got a little lost in the table helps.
I've got a data.frame I would like to summarize as a (and pardon for the
lack of correct vernacular) data collection matrix. My data looks like,
stand siteindex age acres pct.acres
1232 116 45 8477.3105
f-as.factor(c(1,1,2,2,3,3,3))
model.matrix(~f-1)
Patrik Waldmann Patrik.Waldmann at genfys.slu.se writes:
:
: Hello again,
:
: I was too quick before. What I was looking for was a function that
: constructs the design (or incidence) matrix (X in a linear model) from a
: factor. Uwe Ligges
Patrik Waldmann [EMAIL PROTECTED] writes:
Hello again,
I was too quick before. What I was looking for was a function that
constructs the design (or incidence) matrix (X in a linear model) from a
factor. Uwe Ligges suggested using model.matrix and this does almost what I
want, but it is
I don't think the design matrix and the incidence matrix are the same thing.
A design matrix is the X matrix in a usual linear model, which would
(usually) include the column of ones, and the contrasts. So model.matrix()
is the right thing for a design matrix.
For incidence matrix, you can try:
On Wed, 12 May 2004, Patrik Waldmann wrote:
I was too quick before. What I was looking for was a function that
constructs the design (or incidence) matrix (X in a linear model) from a
factor. Uwe Ligges suggested using model.matrix and this does almost what I
want, but it is first necessary
From: Rolf Turner
Anne Piotet wrote:
What R functionnalities are there to do missing values imputation
(substantial proportion of missing data)? I would prefer to use
maximum likelihood methods ; is the EM algorithm implemented? in
which package?
The so-called ``EM
On 12-May-04 Rolf Turner wrote:
Anne Piotet wrote:
What R functionnalities are there to do missing values imputation
(substantial proportion of missing data)? I would prefer to use
maximum likelihood methods ; is the EM algorithm implemented? in
which package?
The so-called ``EM
That's not an algorithm. It is a recipe for deriving an algorithm.
algorithm - A detailed sequence of actions to perform to accomplish some
task. Named after an Iranian mathematician, Al-Khawarizmi.
Technically, an algorithm must reach a result after a finite number of
steps, thus
(Ted Harding) [EMAIL PROTECTED] writes:
On 12-May-04 Rolf Turner wrote:
Anne Piotet wrote:
What R functionnalities are there to do missing values imputation
(substantial proportion of missing data)? I would prefer to use
maximum likelihood methods ; is the EM algorithm implemented? in
Picky, picky.
Details are in the eyes of the beholder.
Prof Brian Ripley [EMAIL PROTECTED] writes:
That's not an algorithm. It is a recipe for deriving an algorithm.
algorithm - A detailed sequence of actions to perform to accomplish some
task. Named after an Iranian mathematician,
A.J. Rossini wrote:
Picky, picky.
Details are in the eyes of the beholder.
algorithm - A detailed sequence of actions to perform to accomplish some
task. Named after an Iranian mathematician, Al-Khawarizmi.
Personally I like the first definition of 'Algorism, Algorithm' in the
1913 Websters
Dear Patrick,
Try model.matrix(~ factor - 1).
I hope this helps,
John
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Patrik Waldmann
Sent: Wednesday, May 12, 2004 11:34 AM
To: '[EMAIL PROTECTED]'
Subject: [R] Design matrix not identity
Hello
Hi group,
I am trying to do a RF with approx 250,000
cases. My objective is to determine the risk factors
of a person being readmitted to hospital (response=1)
or else (response=0). Only 10%, or 25,000 cases were
readmitted. I've heard about down-sampling and class
weight approach and am
On 12-May-04 A.J. Rossini wrote:
(Ted Harding) [EMAIL PROTECTED] writes:
[...]
Algorithm, this, or not
[...]
Thanks, Ted :-) -- to extend it a bit, one can imagine the use of
approximate solutions to the 2 steps (simulation methods to get
expected values, similar range of approaches for
Check out aggregate.table in package gregmisc .
Jeff D. Hamann jeff.hamann at forestinformatics.com writes:
:
: I've got a newbie question and I got a little lost in the table helps.
: I've got a data.frame I would like to summarize as a (and pardon for the
: lack of correct vernacular) data
-Original Message-
From: Prof Brian Ripley [mailto:[EMAIL PROTECTED]
Sent: Tuesday, May 11, 2004 11:21 PM
To: Vadim Ogranovich
Cc: R-Help
Subject: Re: [R] recover should send messages to stderr, not stdout
...
Note that some of us consider recover() to be designed for
Dear R-helpers,
I would like to change the default value of a functions that is call by a
package. The function deparse() is called by a set of functions from a
library and because I don't think is appropriate to change the library, I am
trying to change the default value in deparse().
On Wed, 12 May 2004, Vadim Ogranovich wrote:
-Original Message-
From: Prof Brian Ripley [mailto:[EMAIL PROTECTED]
Sent: Tuesday, May 11, 2004 11:21 PM
To: Vadim Ogranovich
Cc: R-Help
Subject: Re: [R] recover should send messages to stderr, not stdout
...
Note
Thanks Brian.
The EM algorithm requires an ``E'' step and an ``M'' step. Harding
and Rossini appear to be seriously suggesting that an R function
could be written which would
(a) Perform the E step in arbitrary contexts, and
(b) For that given expected value, work out a
Rolf Turner [EMAIL PROTECTED] writes:
The EM algorithm requires an ``E'' step and an ``M'' step. Harding
and Rossini appear to be seriously suggesting that an R function
could be written which would
(a) Perform the E step in arbitrary contexts, and
(b) For that given expected
Breiman Cutler's version 5 of the Fortran code implements a weighting
scheme that is more effective than the old classwt. Basically:
1. class weights are used in computing the Gini index.
2. At terminal nodes, weighted votes are taken to determine the prediction
for the node.
3. Average weights
Is there a way I can have R automatically execute the commands in a source file
without ever having to use R interactively? If so, what arguments should I pass to the
UNIX call to R? I need to do this to run several R jobs in parallel.
An alternative may be to have R and an R package behave as
Bickel, David [EMAIL PROTECTED] writes:
Is there a way I can have R automatically execute the commands in a
source file without ever having to use R interactively? If so, what
arguments should I pass to the UNIX call to R? I need to do this to
run several R jobs in parallel.
Yes; help(BATCH)
hi: I have a y vector and an x data frame. is it possible to use the
standard linear regression (lm, summary.lm) without having to construct
a formula, i.e., all x variables should be used.
x - data.frame( rnorm(10), rnorm(10) ); # -- well, this would
really be read from a file
y -
x - data.frame( x1=rnorm(10), x2=rnorm(10) )
lm(y ~., data=x)
seems to be what you want.
BTW, R is not C/Perl/... and ; is a separator, not a terminator.
On Wed, 12 May 2004, ivo welch wrote:
hi: I have a y vector and an x data frame. is it possible to use the
standard linear regression
lm(y~.,data=x)
ivo welch iaw4 at welch.som.yale.edu writes:
:
: hi: I have a y vector and an x data frame. is it possible to use the
: standard linear regression (lm, summary.lm) without having to construct
: a formula, i.e., all x variables should be used.
:
: x - data.frame(
R-users,
I am simulating a birth process for 4 classes of individuals with
l[i] being the average No. fetuses per individual. However, I need to
bound the resulting values for each generated rpois to be =3 (no
individual can have 3 offspring). I have not been able to figure out how
On 12-May-04 Rolf Turner wrote:
The EM algorithm requires an ``E'' step and an ``M'' step. Harding
and Rossini appear to be seriously suggesting that an R function
could be written which would
(a) Perform the E step in arbitrary contexts, and
(b) For that given expected value,
In S-plus I have depended upon a simple extended ls() function developed by John
wallace called lsx()
Reason: lsx() allows me to search for object names in my local workspace matching an
unquoted character string.
Since I forget all the function names or data object names I have added, I can
Tim,
On the configure problem I've determined that this is a bug in the file
acinclude.m4 provided in the R distribution. The test function
R_SYS_POSIX_LEAPSECONDS needs to have ct initialized. That it works on
other systems is fortuitous. Can you provide this feedback to the
developers?
Bret,
the following should do:
x - sum(pmin(rpois(recruit.f, l),3))
Giovanni
Date: Wed, 12 May 2004 16:11:08 -0500
From: Bret Collier [EMAIL PROTECTED]
Sender: [EMAIL PROTECTED]
Precedence: list
R-users,
I am simulating a birth process for 4 classes of individuals with
l[i]
Dear Sir or Madam,
What is the formula for power.t.test(delta=delta, sd=segma, sig.level=0.05,
power=0.8, type=one.sample, alternative=one.sided)$n?
Thank you very much for the help!
Best,
Xiaorong
__
[EMAIL PROTECTED] mailing list
Hi
I'm using lme4 to do random effects modelling.
I keep getting the following error message:
Error in EMsteps-(*tmp*', value = control) :
invalid source matrix
I get the error when I include more than one random effect in the model, sometime I'm
able to get two. I've looked into
On 12-May-04 Bret Collier wrote:
R-users,
I am simulating a birth process for 4 classes of individuals
with l[i] being the average No. fetuses per individual. However, I
need to bound the resulting values for each generated rpois to be =3
(no individual can have 3 offspring). I
You can do that with ls and the pattern= argument. For example,
ls(pattern = my)
finds all objects with my anywhere in the name. See ?ls
Andrew White andrew_white at hmsa.com writes:
:
: In S-plus I have depended upon a simple extended ls() function developed by
John wallace called lsx()
On Wednesday 12 May 2004 16:48, Matt Loveland wrote:
Hi
I'm using lme4 to do random effects modelling.
I keep getting the following error message:
Error in EMsteps-(*tmp*', value = control) :
invalid source matrix
I get the error when I include more than one random effect in
I am attempting to produce a bar chart and am having some trouble with the
panel.barchart command. The definition of X is that it's the proper extent of the
bar. Is that the maximum value? Or is it the labels for my vertical axis, since the
axes seem to be switched for this command? Is there a
Perhaps you really prefer something with a continuous first derivative?
In that case, all the continuous cumulative distribution functions have
a sigmoidal shape and might be suitable. You could fit pnorm, plogis or tanh
with suitable scaling and location parameters using nls. An example
of
When I tried to read all the entries (after searching the FAQ) for row
names xtable, I get
START-INFO-DIR-ENTRY * R FAQ: (R-FAQ). The R statistical system ...
... xtable* Export data to LaTeX and HTML tables. ... For dropping the row
names of a matrix
`x', it may be easier to use `rownames(x) -
Hello, everybody,
I met a big problem. What I want is to use c code in R. I have load a test
code conv.c in R and run it successfully. However, in my real code, I use
matlab c library (Matrix Inverse function). After I use command R CMD SHLIB
to share library, I can not use dyn.load to
Gabor Grothendieck [EMAIL PROTECTED] writes:
Perhaps you really prefer something with a continuous first derivative?
In that case, all the continuous cumulative distribution functions have
a sigmoidal shape and might be suitable. You could fit pnorm, plogis or tanh
with suitable scaling and
On Wed, 12 May 2004 20:13:09 +0100, Jesus Frias [EMAIL PROTECTED]
wrote:
Dear R-helpers,
I would like to change the default value of a functions that is call by a
package. The function deparse() is called by a set of functions from a
library and because I don't think is appropriate to
Don't know about xtable but here is an alternative:
require(Hmisc)
data(iris)
iris4 - iris[1:4,] # first 4 rows of iris as test
latex(iris4,file=) # with row names
attr(iris4,row.names) - NULL
latex(iris4,file=) # without row names
Jeff D. Hamann jeff.hamann at forestinformatics.com writes:
:
On Tue, 11 May 2004, Rui wrote:
Hi all,
I want to know some details about the c routine lasso in the functions
of gl1ce() . However, I have following troubles. First, I can not
find the routine in the local directories of this function (or package).
Then something is wrong with your
recover() is a pure R function, so you can very easily edit it to meet
your requirements. You could also submit a patched version for
consideration after testing.
Note that some of us consider recover() to be designed for
interactive-only use, and use something like
I've recently updated nlrq to get rid of these warnings. Peter's diagnosis
of the optim problem seems entirely consistent with the evidence: my mac
running R-devel was Jim's source on the error free case. And my solaris
machine (running 1.9.0 unpatched) does eventually have a problem with
optim:
On Tue, May 11, 2004 at 07:10:36AM -0700, Thomas Lumley wrote:
On Mon, 10 May 2004, Paul Johnson wrote:
Dear Everybody:
I'm doing my usual how does that work in R thing with some Stata
projects. I find a gross gap between the Stata and R in Cox PH models,
and I hope you can give me
On Wed, 12 May 2004, Martin Kerick wrote:
I have a list-object RGList named RG containing two vectors (RG$R and RG$G)
and a data.frame(RG$genes) with 8 variables. Each of the variables of the
data.frame have the same length as the two vectors. I think the data in the
RGList is structured in
Hello all!
I fitted a variogram model using GeoR to regression model residuals. I would
like to assess how much of the variation
is explained by the non-spatial regression model and the spatial model. Ive
got the log likelihood, AIC and deviance respectively residual SS and R2 for
the
Hi All,
I'm building a package using C/C++ and Fortran code which usually runs fine.
However I do get occasional bus errors around the time of exiting one of the
C functions.
Where do I need to be looking to solve this problem?
Do bus errors stem from unmapped memory exceptions? How can I find
Hi!
After loading some functions in R1.9.0 and some computations I am getting the
following errors (see below with plot).
I have now idea how to figure out whats wrong. Can anyone give me a hint where to
start to debug?
Eryk
plot(1,1) fine.
plot(hist(1:10,plot=F))
Error in xy.coords(x, y,
Dear All,
Im trying to solve a two variable Quadratic Program, is there a way to
solve the following problem with quadpro in R :
Min(K) Max(alfa){ C= alfa =0;t(alfa)%*%y=0}
t(e)%*% alfa - 0.5t(alfa)%*% D %*%alfa
subject to
trace(K) = c
K = \sum_{i=1}^m mu_i K_i
K = 0
witek wrote:
Hi!
After loading some functions in R1.9.0 and some computations I am getting the
following errors (see below with plot).
I have now idea how to figure out whats wrong. Can anyone give me a hint where to
start to debug?
Eryk
plot(1,1) fine.
plot(hist(1:10,plot=F))
Error in
On Wed, May 12, 2004 at 12:19:41PM +0200, witek wrote:
Hi!
After loading some functions in R1.9.0 and some computations I am getting the
following errors (see below with plot).
I have now idea how to figure out whats wrong. Can anyone give me a hint where to
start to debug?
Eryk
Göran Broström [EMAIL PROTECTED] writes:
coef exp(coef) se(coef) zp
I(haz.wst * 1e-06) 0.08479 1.090.095 0.8920 0.37
pol.free 0.00896 1.010.170 0.0526 0.96
Likelihood ratio test=0.76 on 2 df, p=0.685 n= 21
Dear R list
I have an m * n matrix P and a vector V of length n containing indices for
rows in P.
For each of the m columns I want to extract the value in the row specified
by V, and put these values into a new vector W of length n.
At present I am doing this with a for
loop, but I imagine there
Dear Dan,
you could just use the following:
V - cbind(V, 1:n)
W - P[V]
I hope this helps.
Best,
Dimitris
Dimitris Rizopoulos
Doctoral Student
Biostatistical Centre
School of Public Health
Catholic University of Leuven
Address: Kapucijnenvoer 35, Leuven, Belgium
Tel: +32/16/396887
Fax:
A very brief question.
I have managed to put greek letters in my axis labels using
expression().
Question is, is there an easier (shorter) way than typing
plot(runif(10),ylab=expression(NH[4]~(mu~M)),xlab=Time (h))
to get the subscripts and micro sign.
Like NH$_4$ $\mu M$ would be nice :)
Daniel Bebber wrote:
Dear R list
I have an m * n matrix P and a vector V of length n containing indices for
rows in P.
For each of the m columns I want to extract the value in the row specified
by V, and put these values into a new vector W of length n.
At present I am doing this with a for
Andersson, Henrik wrote:
A very brief question.
I have managed to put greek letters in my axis labels using
expression().
Question is, is there an easier (shorter) way than typing
plot(runif(10),ylab=expression(NH[4]~(mu~M)),xlab=Time (h))
to get the subscripts and micro sign.
Like NH$_4$
Hi,
resending this question; guess someone missed it (?)
Thanks,
Maarten
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Hoeven, Maarten
van der
Sent: Monday, May 10, 2004 12:34 PM
To: [EMAIL PROTECTED]
Subject: [R] sqlSave with underscores in table
88 matches
Mail list logo