Hi everybody,
I'm trying to use cross-validation (cv.glm) for count data. Does someone know
which is the appropriate cost function for Poisson distribution?
Thank you in advance.
Valerio.
Conservation Biology Unit
Department of Environmental and Territory Sciences
University of Milano-Bicocca
Dear Michael
It calls the function pt() to calculate the p-value based on
your t statistics and the degree of freedoms.
If you are interested how it is calculated in details, you can
have a look into the source code:
/R-2.4.0/src/nmath/pt.c
There you can find the C-code.
Hope this helps
On Wed, 15 Nov 2006, [EMAIL PROTECTED] wrote:
I'm trying to use cross-validation (cv.glm) for count data. Does someone
know which is the appropriate cost function for Poisson distribution?
It depends on the scientific problem, not the distribution.
You could use the deviance but it may well
Mark,
I think you should realize that SQL is evolved from a different culture, large
databases. It does not have these fancy data types that we find so usefull in
R, so somewhere there is guessing what class a variable is.
I tried to examine what NULL values would do on my computer Linux, Suse
Hi everyone,
I am fitting a bivariate smoothing model by using gam.
But I got an error message like this:
Error in eigen(hess1, symmetric = TRUE) : 0 x 0 matrix
- this is a known problem in mgcv 1.3-20 (an optimizer fails to cope with
convergence in one step). It's fixed in 1.3-21, which
chao gai wrote:
Why would the database be so versatile anyway. Inf, -Inf, NA and NaN are not
exaclty what one would use for say an inventory, personel adminstration or
the such. These are probably not part of the SQL standard anyway.
Thanks for your reply. I have to agree with most of your
Dear All:
I used optim() to minimise the loglikelihood function for fitting data to
negative binomial distribution. But there initial value of log-likelihood and
iteration 10 value are reasonable. for example:
initial value 1451657.994524
iter 10 value 47297.534905
iter 20 value
Hi, dear all
As Professor LUMLEY suggest, I use the withReplicates function to comute SE of
a complex statistics here below the source I wrote :
wi-weights(EC2deg)
Qratio-function(w=wi,don=EC2,x,a1=.2,a2=.8){
Hi,
i have many variables for in example 4weeks and want to do
aggregations, like mean standard , deviation etc..
With mean it works but how i can calculate the standard deviation for
the 4weeks and for every ID.
many thanks regards, christian
week1 - grep((_PRO_001),names(dmx3),perl=T)
Hello,
I actually have an understanding problem about estimating ARIMA-Models.
I use the package forecast
and here I simply do best.arima.
I can get fitted values by the comand fitted()
To estimate the forecasterrors
x=c(1:20)
y=best.arima(x)
fitted(y)
forecasterrors(fitted(y), x)
My
Hello all,
I want to draw a grid behind my graphes, using lattice package.
I manage to do it with instructions like this one :
xyplot(Sepal.Length + Sepal.Width ~ Petal.Length ,
data = iris, allow.multiple = TRUE, scales = same,type=l,
panel = function(...) { panel.grid(h=-1, v=
R-help,
I want to remove the following strings
cpue and nogd
string - c(upsanogd ,toskanogd , hysunogd , konganogd
,gullaksnogd , longunogd , blalongunogd , brosmunogd)
I could use first : first - gsub(cpue , , string)
and then : second - gsub(nogd , , first)
Can it be done at
I'm very new to R, so please forgive me if I just missed the answer in
existing documentation...
I have a data set with at least three columns, X, Y, and Z.
I want to produce a chart where one axis shows all the unique values of X,
and the other axis shows all the unique values of Y. Each cell
1. Another way to address this is to include g in the type vector:
xyplot(Sepal.Length + Sepal.Width ~ Petal.Length ,
data = iris, allow.multiple = TRUE, scales = same, type = c(l, g))
2. Also look in the example section of:
library(zoo)
?xyplot.zoo
which gives an example of plotting a
On 11/15/2006 8:29 AM, Luis Ridao Cruz wrote:
R-help,
I want to remove the following strings
cpue and nogd
string - c(upsanogd ,toskanogd , hysunogd , konganogd
,gullaksnogd , longunogd , blalongunogd , brosmunogd)
I could use first : first - gsub(cpue , , string)
and
Dear Michael,
This looks like a multivariate simple regression -- that is, 12 response
variables, one predictor. If the data are in the matrix X, then lm(X[,1:12]
~ X[,13]) should do the trick.
I hope this helps,
John
John Fox
Department of Sociology
McMaster
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
A new package named BiasedUrn is available.
This package implements various noncentral hypergeometric distributions,
univariate and multivariate.
Includes the distribution you get when taking colored balls from an urn
without replacement, with bias.
These distributions have many applications in
Dear all:
I got some error messages when I installed R-2.4.0 on Compaq Tru64 Unix
V5.1B, I set the R_HOME=/shs/R and then try 'make' command after
configuration and it showed Make: Cannot open /share/make/vars.mk.
Stop.,after that,I used the 'gmake' to implement.
The following is my
One way is to graft the stratified sampling code from the classification
part onto the regression part. I will get to it eventually, but just
not now.
Andy
From: Naiara Pinto
Dear all,
I am doing a regression in ramdomForest, using the option
sampsize reduce the number of records used
Thank you, Spencer.
It *does* depend on the exact environment such as timezone,
locale, ???
In my (Linux) setup, I can reproduce Rogers' 1-off problem
using exactly your (Spencer's) R script.
Carefully considering the warning in ?POSIXlt,
and using Sys.getenv(TZ) to see that I had no timezone
Christian Convey wrote:
I'm very new to R, so please forgive me if I just missed the answer in
existing documentation...
I have a data set with at least three columns, X, Y, and Z.
I want to produce a chart where one axis shows all the unique values of X,
and the other axis shows all the
I'm not sure I understand the question, but you might look into the
following functions:
unique
heatmap
image
Again, if I understand the question, you would create a length(unique
(x)) by length(unique(y)) sized matrix, and fill it with appropriate
values of z. Then pass that to heatmap or
Hi all,
Does anyone know what happened to the dna library or the dotmatrix
function? For the life of me, I can't find it anywhere with the
exception of this reference:
http://rss.acs.unt.edu/Rdoc/library/dna/html/dotmatrix.html
Thanks!
Jeff.
http://www.nd.edu/~jspies/
On Tue, 14 Nov 2006, RMan54 wrote:
This works for the original posted question:
n-5
title - bquote(bold(paste(Figure , .(n), : Plot , C[max], versus
CrCl)))
plot(1, main=title)
However, my problem is that I want to define the text before the value of n
is known.
The idea is that the
I'm trying to build R 2.4.0 on an IBM P5-570 that's running
AIX 5.3. I'm using xlc 7.0, xlc++ 7.0 and xlf 9.1 in 32 bit
mode (OBJECT_MODE is 32 in the build environment).
The source is the patched version of 2.4.0, downloaded yesterday.
Configure options were --prefix, --srcdir, --x-includes,
Hello!
I have some data stored into 2 separate csv file. 1 file (called A.csv) (12
results named Group1, Group2, Group3, etc...) odds ratios, 2 file (called
B.csv) 12 corresponded errors.
How to import that data into R and make forest plot like I saw inside help file
Rmeta and meta with
If you know how to use the unix/linux tools already, then you may want
to look at cygwin (http://www.cygwin.com/), it allows those of us
trapped in a windows world to still lead productive lives with the
unix/linux tools.
--
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain
Hi Rogerio:
Here is what I tried. I have used only data points 1 to 1447 but the same
problems appear if I used all the data in the file I sent .
library(waveslim)
infile -C:\\airon.csv
ckhdat - read.csv(infile,header=TRUE, quote=)
...
1464
Is this what you want? :
gsub(cpue\|nogd, , string)
John
---
Web sites:
www.ifr.ac.uk
www.foodandhealthnetwork.com
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Luis Ridao Cruz
Sent: 15 November 2006 13:29
To: r-help@stat.math.ethz.ch
Jeffrey Robert Spies [EMAIL PROTECTED] writes:
Hi all,
Does anyone know what happened to the dna library or the dotmatrix
function? For the life of me, I can't find it anywhere with the
exception of this reference:
http://rss.acs.unt.edu/Rdoc/library/dna/html/dotmatrix.html
Erin == Erin Hodgess [EMAIL PROTECTED]
on Tue, 14 Nov 2006 17:30:11 -0600 writes:
Erin Hi R People:
Erin If you have a linear programming problem in which
Erin some of the constraints have the =, some
Erin have = and some have =, all in the same problem,
Erin should the
I have two variables, minutereturnsa which can be thought of as my
independent variable and minutereturnsb which can be thought of as my
dependent variable. When I run correlations on the two variables,
depending on which of the three methods I use, I get values of between
-.15 through -.19.
I am trying to plot the time series decomposition using plot(stl..).
Eventhough I understand why the scale of the 4 plots is better to be
unequal, I would like to plot all 4 in the same scale (otherwise
interpretation at a simple look may be misleading). Is there a way I could
do so (easier
Hi
is there an easy way/ R-function to calculate the numerical maximum
likelihood estimators for a Student's t-distribution?
I searched the mailing list archive the last 30mins but didn't find an answer.
Regards
Ben
__
R-help@stat.math.ethz.ch mailing
Apologies if this is the wrong list, but could somebody put the
information on how to create the graphs on
http://www.r-project.org/screenshots/screenshots.html (or a link to
these instructions) next to the graphs?
__
R-help@stat.math.ethz.ch mailing
oops, I forgot to square !! Thanks Chuck. I would have
spent the rest of the day and then some trying to figure that one out
-Original Message-
From: Chuck Cleland [mailto:[EMAIL PROTECTED]
Sent: Wednesday, November 15, 2006 11:46 AM
To: Leeds, Mark (IED)
Subject: Re:
On Wed, 15 Nov 2006, Benjamin Dickgiesser wrote:
Hi
is there an easy way/ R-function to calculate the numerical maximum
likelihood estimators for a Student's t-distribution?
I searched the mailing list archive the last 30mins but didn't find an answer.
See fitdistr() in MASS.
MLE of what,
Hi,
I have encountered this problem quite a few times and thought I would
ask.
Let's say that I have two endpoints, a and b, which are integers. If
a = b, I would like to get a:b, but if a b, then numeric(0), for
example:
myseq(3,5) = 3:5
myseq(3,3) = 3
myseq(3,2) = numeric(0)
The operator :
I need to estimate all parameters (except maybe df). Thank you for
pointing me into a direction, I will have a look.
The aim is to use a fat-tail distribution to calculate the Value At
Risk instead of using the Normal distribution.
Ben
On 11/15/06, Prof Brian Ripley [EMAIL PROTECTED] wrote:
On
Dear R users,
I am trying to reproduce table 7.4.12 (page 131) from Snedechor
Cochran (eigth edition); the example is counts of weed seeds with a
fitted Poisson distribution, tested for goodness-of-fit using a Chi-square:
observed=c(3,17,26,16,18,9,3,5,0,1,0,0)
Dear R users,
I am trying to reproduce table 7.4.12 (page 131) from Snedechor
Cochran (eigth edition); the example is counts of weed seeds with a
fitted Poisson distribution, tested for goodness-of-fit using a Chi-square:
observed=c(3,17,26,16,18,9,3,5,0,1,0,0)
On Wed, 2006-11-15 at 11:35 -0500, Tamas K Papp wrote:
Hi,
I have encountered this problem quite a few times and thought I would
ask.
Let's say that I have two endpoints, a and b, which are integers. If
a = b, I would like to get a:b, but if a b, then numeric(0), for
example:
Knut M. Wittkowski wrote:
Apologies if this is the wrong list, but could somebody put the
information on how to create the graphs on
http://www.r-project.org/screenshots/screenshots.html (or a link to
these instructions) next to the graphs?
A nice reference for graphics with R can be found
Hi,
I would like to get the piecewise polynomial form of a B-spline basis
(for repeated calculations of the same spline). polySpline gives the
polynomial representation of a particular spline. I wonder if anybody
has code to do that for the basis itself...
Thanks,
Tamas
On Wed, 15 Nov 2006, Tamas K Papp wrote:
Hi,
I have encountered this problem quite a few times and thought I would
ask.
Let's say that I have two endpoints, a and b, which are integers. If
a = b, I would like to get a:b, but if a b, then numeric(0), for
example:
myseq(3,5) = 3:5
... seq(a, b, length = ifelse(a = b, b - a + 1, 0))
seq(a,b,length = max(0,b-a+1))
seems a bit simpler and more transparent.
-- Bert
Bert Gunter
Genentech Nonclinical Statistics
South San Francisco, CA 94404
__
R-help@stat.math.ethz.ch mailing list
Hello,
I work on large matrices and found something interesting. For multiplication of
matrices, the order has a huge influence on computing time when one of them is
a sparse matrix. In the below example, M is a full matrix and A is a sparse
matrix in a regular matrix class. A %*% M takes
Christoph Scherber [EMAIL PROTECTED] writes:
Dear R users,
I am trying to reproduce table 7.4.12 (page 131) from Snedechor
Cochran (eigth edition); the example is counts of weed seeds with a
fitted Poisson distribution, tested for goodness-of-fit using a Chi-square:
Hi,
I am sure this has come up before, but my searches of the archive
didn't give any results (maybe I didn't use the right keywords, but if
I use too many, the search times out).
I have a vector of dimensions n, length is not fixed, eg
n - c(4,5,7)
or
n - c(19,4,5,7)
and a function f that
On Wed, Nov 15, 2006 at 01:22:19PM -0500, YONGWAN CHUN wrote:
I work on large matrices and found something interesting. For multiplication
of matrices, the order has a huge influence on computing time when one of
them is a sparse matrix. In the below example, M is a full matrix and A is a
I finally found the answer: the easiest way to enforce this is just
removing the 2 knots, like this:
## thanks for all those who replied to the increasing sequences questions
incseq - function(a,b) {
if (a = b)
seq.int(from=a,to=b)
else
integer(0)
}
## remove augment knots
On 11/15/06, YONGWAN CHUN [EMAIL PROTECTED] wrote:
I work on large matrices and found something interesting. For multiplication
of matrices, the order has a huge influence on computing time when one of
them is a sparse matrix. In the below example, M is a full matrix and A is a
sparse
There are 12 response variables, columns 1 to 12 are response variables, i.e.,
these are y's, they all regress to the 13th column, which is the predictor,
i.e. the X.
Let's take column 1, call this Y1, and there are n rows(n samples) of it,
I need Y1= b0_1 + b1_1* X + epsilon, where X is the
Thanks, let me try to clarify my question with an example.
Suppose I have the following data:
Gender, Major, Course-Grade
F, Psy, 3.5
F, Psy, 3.1
M, Hst, 3.7
F, Hst, 3.6
M, Hst, 2.6
M, Eng, 3.9
I want to compute a table like the following:
X-axis: Gender
Y-axis: Major
Cell(x,y) = mean
On Wed, 2006-11-15 at 15:03 -0500, Christian Convey wrote:
Thanks, let me try to clarify my question with an example.
Suppose I have the following data:
Gender, Major, Course-Grade
F, Psy, 3.5
F, Psy, 3.1
M, Hst, 3.7
F, Hst, 3.6
M, Hst, 2.6
M, Eng, 3.9
I want to compute a
Marc's solution looks a bit easier but here are a few more anyways:
# 1
reshape(DF, dir = wide, timevar = Gender, idvar = Major)
# 2
library(reshape)
DFm - melt(DF, id = 1:2)
cast(DFm, Major ~ Gender, fun = mean)
On 11/15/06, Christian Convey [EMAIL PROTECTED] wrote:
Thanks, let me try to
Airon,
I don't think you have to find an English computer 'cause the
following must work in your Chinese one :-)
Let me explain. First of all, change your lines to
xdata - ckhdat$Adj..Close[1:1447]
#names(ckhdwt.la8) - c(w1, w2, w3, w4, w5,w6, v6)
note the # sign, i.e., DO NOT change the
That did it! Thanks Marc.
- Christian
On 11/15/06, Marc Schwartz [EMAIL PROTECTED] wrote:
On Wed, 2006-11-15 at 15:03 -0500, Christian Convey wrote:
Thanks, let me try to clarify my question with an example.
Suppose I have the following data:
Gender, Major, Course-Grade
F, Psy,
I need to run a regression analysis with a large number of samples. Each
sample (identified in the first file column) has its own x and y values. I
will use the same model in all samples. How can I run the model for each
sample? In SAS code I would use the BY SAMPLE statement.
Alvaro
Hi Kevin,
sadly, you have (again) not provided enough information to enable us
to make a constructive response - for example, what package are you
using? I assume that it's not mgcv, as mgcv does have a predict.gam
with an se.fit argument:
predict.gam package:mgcv R
Xin jasonshi510 at hotmail.com writes:
Dear All:
I used optim() to minimise the loglikelihood function for fitting data to
negative binomial
distribution. But there initial value of log-likelihood and iteration 10 value
are reasonable. for example:
initial value 1451657.994524
iter
Hi, All:
Does 'nlme' require the nonlinear function to be differentiable?
I'm fitting structural change models to many related time series, and
I'd like to improve the estimates of the change point times through some
type of pooling. Unfortunately, I've so far been unable to get
To all:
We have an opening for a recent graduate to work as a Statistical
Programmer at Advanced Micro Devices (AMD) in Sunnyvale CA beginning
January 2007. AMD is a global supplier of microprocessors and
silicon-based solutions to the communications and computer industries.
This position will
How about I make a design matrix as follows:
1 d1,1 0 0 0 0 0 0 0
00 1 d1,2 0 0 0 0 0
00 0 0 1 d1,3 0 0 0
...
...
000 0 0 0 ... ... 1 d1,12
The above matrix will work for the 1st row of Y data;
d1, 1 means the 1st
Dear List,
Does anyone have experience using mmlcr Package? I am working on a
longitudinal dataset. The model fitting is fine. However, I had a problem
with the command plot.mmlcr. It seems to me it can not plot a very
nice trajectory plot for longitudinal dataset. I wonder whether somebody can
Hi Rogerio:
Thks a lot. It works.
By the way, I have 2 related sides issues that need some help:
(1) What I want to do is this
- do modwt on original time series
- do thresholding on wavelet coefficients
- obtain the inversed smoothed and detailed components of the original
On 11/14/06, Ravi Varadhan [EMAIL PROTECTED] wrote:
I am trying to do the following computation:
p - rep(0, n)
coef - runif(K+1)
U - matrix(runif(n*(2*K+1)), n, 2*K+1)
for (i in 0:K){
for (j in 0:K){
p - p + coef[i+1]* coef[j+1] * U[,i+j+1]
} }
___
Hello,
Not sure if this counts as a bug or not, but I just noticed in R 2.4.0
that : and interaction are not quite equivalent. For example:
x - factor(letters[1:4])
y - factor(letters[11:14])
x:y
[1]
Hi, I'm just starting out on R. I have an example data frame listed below.
I have regressed Y on X1. lm(Y~X1). Would like to obtain the Brown-Forsythe
test on residuals.
I am unsure of how to conduct the following steps.
(1) Inputting the residuals from regression into the data frame.
(2)
Dear Prof. Ripley,
Finally, I succeeded in building R from source. I just changed WINHELP part in
'MkRules' file from WINHELP = CHM to WINHELP = NO. Then R was built
successfully.
Thank you so much.
Yongwan Chun
- Original Message -
From: Prof Brian Ripley [EMAIL PROTECTED]
Date:
On Wed, 15 Nov 2006, Michael wrote:
How about I make a design matrix as follows:
1 d1,1 0 0 0 0 0 0 0
00 1 d1,2 0 0 0 0 0
00 0 0 1 d1,3 0 0 0
...
...
000 0 0 0 ... ... 1 d1,12
The above matrix will work for the
Greetings:
I've installed Rtools, MikTeX, perl, minGW, and HTML Help Workshop, and have
succeeded in making, checking (using R CMD check mypkg) then building the
simple example package.skeleton(list=c(f,g,d,e), name=mypkg) R
CMD build mypkg produces a tarball. I don't know how to get a zip
How do you build your packages?
have you tried
R CMD build --binary mypkg
b
On Nov 16, 2006, at 12:32 AM, Charles Annis, P.E. wrote:
Greetings:
I've installed Rtools, MikTeX, perl, minGW, and HTML Help Workshop,
and have
succeeded in making, checking (using R CMD check mypkg) then
Another way is
reg-lapply(split(data,data$byvar),lm,formula=y~x)
Justin BEM
Elève Ingénieur Statisticien Economiste
BP 294 Yaoundé.
Tél (00237)9597295.
- Message d'origine
De : Chuck Cleland [EMAIL PROTECTED]
À : Alvaro [EMAIL PROTECTED]
Cc : r-help@stat.math.ethz.ch
Envoyé le :
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Peter Dalgaard
Sent: Wednesday, November 15, 2006 5:58 PM
To: Jeffrey Robert Spies
Cc: r-help
Subject: Re: [R] Dotmatrix Plots
Jeffrey Robert Spies [EMAIL PROTECTED] writes:
Hi all,
Does
If you want a quick way to do this that mean you have a way to do that who is
not to fast. First propose it.
Justin BEM
Elève Ingénieur Statisticien Economiste
BP 294 Yaoundé.
Tél (00237)9597295.
- Message d'origine
De : Louis Tay [EMAIL PROTECTED]
À : r-help@stat.math.ethz.ch
On Thu, 16 Nov 2006, Airon Yiu wrote:
Hi Rogerio:
Thks a lot. It works.
By the way, I have 2 related sides issues that need some help:
(1) What I want to do is this
- do modwt on original time series
- do thresholding on wavelet coefficients
- obtain the inversed smoothed and
78 matches
Mail list logo