Dear list,
I'm trying to understand what kind of uncertainty intervals are
plotted for survival::survfit.coxph when 'conf.int=TRUE' is called,
e.g.:
f <- coxph(Surv(futime, fustat) ~ age, data = ovarian)
s <- survfit(f, newdata=data.frame(age=c(30, 60)))
plot(s, conf.int=TRUE)
Are these bands th
Dear list,
I'm trying to understand what kind of uncertainty intervals are
plotted for survival::survfit.coxph when 'conf.int=TRUE' is called,
e.g.:
f <- coxph(Surv(futime, fustat) ~ age, data = ovarian)
s <- survfit(f, newdata=data.frame(age=c(30, 60)))
plot(s, conf.int=TRUE)
Are these bands th
On Wed, Jan 20, 2010 at 8:18 AM, Douglas Bates wrote:
> On Tue, Jan 19, 2010 at 9:26 AM, Martin Maechler
> wrote:
>> Scanning for 'Matrix' in old R-help e-mail, I found
>>
>>>>>>> "GA" == Gad Abraham
>>>>>>> o
On Wed, Jan 20, 2010 at 2:26 AM, Martin Maechler
wrote:
> Scanning for 'Matrix' in old R-help e-mail, I found
>
>>>>>> "GA" == Gad Abraham
>>>>>> on Fri, 27 Nov 2009 13:45:00 +1100 writes:
>
> GA> Hi,
> GA>
On Fri, Nov 27, 2009 at 2:22 PM, David Winsemius wrote:
>
> On Nov 26, 2009, at 9:45 PM, Gad Abraham wrote:
>
>> Hi,
>>
>> I'd like to store large covariance matrices using Matrix classes.
>>
>> dsyMatrix seems like the right one, but I want to
: length of x slot != prod(Dim)
Is there an easier way of doing this?
Thanks,
Gad
--
Gad Abraham
PhD Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~gabraham
_
likely),
you could just run R as normal (single threaded), or do parallel
processing using packages like Rmpi, snow, multicore, etc.. If you use
MPI, you'll probably need a bit of help setting it up properly.
For starters, see
http://cran.r-project.org/web/views/HighPerformanceComputing.html and
the
ecdf,
but perhaps someone can shed some light on it.
Thanks,
Gad
--
Gad Abraham
MEng Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~gabraham
__
graph_1.22.2annotate_1.20.1
[4] xtable_1.5-5AnnotationDbi_1.4.1 Biobase_2.2.1
loaded via a namespace (and not attached):
[1] DBI_0.2-4 RSQLite_0.7-1 XML_2.3-0
--
Gad Abraham
MEng Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Aus
e it helps.
--
Gad Abraham
MEng Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~gabraham
__
R-help@r-project.org mailing list
https://st
ar4 ar5
-0.0483000 0.1355
s.e. 0.1010000 0.1031
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~gabraham
___
ut in the
package glmnet.
It's probably best to chase it up with the package authors since glmnet
is not a base R package.
--
Gad Abraham
MEng Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
g) for
security related issues and/or transient Internet connectivity errors.
It was actually Rolf that got the error message - all I got was the
title "The R and Science..." and nothing under it. Now I get
"LectureMaker.com recommends Flash...". I'm using Flash 9.0.
menu says I've got Mozilla 5.0, Firefox 2.0.0.2
--- whatever that means.
Bottom line --- I can't watch the video.
Firefox 3.0.8 on OS X doesn't work either.
But you can go directly to
http://www.lecturemaker.com/wp-content/uploads/2009/02/lmpremovie.swf
and it will work.
Santosh wrote:
Dear useRs...
How do you all do binning of an independent variable using R? For example,
observations of a dependent variable at times different from a nominal time.
Are there any R functions that help with binning?
Have a look at ?cut.
--
Gad Abraham
MEng Student, Dept
variables with not enough data.
Cheers,
Gad
--
Gad Abraham
MEng Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~gabraham
__
R-help@r-projec
issues of scaling (i.e.,
your variables are very small, use scale() to standardise them.)
--
Gad Abraham
MEng Student, Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~gabraham
___
ten me on what is in the Hessian matrix if 'hessian = TRUE' with optim?
The documentation indicates that this is a numerically differentiated Hessian.
So I am assuming that some small delta is chosen. How is the delta chosen ?
how about http://en.wikipedia.org/wiki/Hessian_matrix ?
/usr/local/bin:/sw/sbin:/usr/X11/bin:/usr/local/git/bin
JAVA_HOME
/Library/Java/Home
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.u
ave a look at Rserve (http://www.rforge.net/Rserve), I've never used it
but it might be useful to you.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse
() function?
Thanks very much in advance,
Shruthi
I don't know if you can get it directly, but you can look at the code in
plot.acf using getAnywhere(plot.acf) to see how it's calculated.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Austr
OD + SESSO + fonte_sct + donor + :
X matrix deemed to be singular; variable 10 14 15 18 20 21 22 24 25 26
why??? anyone know the solution of this??
Perhaps your some of your variables are very highly correlated with each
other, leading to a singular covariance matrix?
--
Gad Abraham
Dept
ot;
Any thoughts about why this seems to now be appearing.
Thanks in advance.
Just a guess, but perhaps you've loaded some other package which
redefines predict() in an incompatible way?
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, A
as test (different test block each time) and
measure the MSE for each repetition. Repeat this for all your time series.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: gabra...@csse.unimelb.edu.au
web: http://www.csse.unimelb.edu.au/~
Daren Tan wrote:
Besides the impute package, are there others that have alternative impute approaches ? I hope to compare their performances.
Have a look at transcan() and impute() in package Hmisc.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria
-validation or the
moving-blocks bootstrap, in essence cutting the timeseries into separate
blocks for training and testing. The blocks have to be large enough to
cover all the ARIMA terms (e.g., more than 10 time steps for an AR(10)).
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourn
cial case of binary outcomes (which is the AUC), Hanley and
McNeil (1982) "The Meaning and Use of the Area under a Receiver
Operating Characteristic (ROC) Curve", Radiology 143:29--36.
Cheers,
Gad
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010
ting/index.html
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.csse.unimelb.edu.au/~gabraham
__
R-help@r-project.org mailing list
https://stat.ethz.ch/mai
n_AU.UTF-8/C/C/en_AU.UTF-8/en_AU.UTF-8
attached base packages:
[1] splines stats graphics grDevices utils datasets methods
base
other attached packages:
[1] Design_2.1-1survival_2.34-1 Hmisc_3.4-3
loaded via a namespace (and not attached):
[1] cluster_1.11.11 grid_2.7.2 latt
Frank E Harrell Jr wrote:
Gad Abraham wrote:
This approach leaves much to be desired. I hope that its
practitioners start gauging it by the mean squared error of predicted
probabilities.
Is the logic here is that low MSE of predicted probabilities equals a
better calibrated model? What
perfect discrimination, but I often find that you can have two
competing models, the first with higher discrimination (AUC) and worse
calibration, and the the second the other way round. Which one is the
better model?
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville
would want the data.
Why not start with some simple plots of the relationships between your
variables? Once you have a feel for the problem, you can look into
modelling it more formally using a suitable regression model.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Park
32 0.7455
X7 32.31 88.59 0.36 0.7153
X8-123.62 322.01 -0.38 0.7011
X9 174.07 464.86 0.37 0.7081
X10-36.59 99.23 -0.37 0.7123
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECT
Y, weights = weights, start = start, etastart =
etastart, :
fitted probabilities numerically 0 or 1 occurred
What's the reason for this discrepancy?
Thanks,
Gad
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTEC
alues related to loadings
in R?
You can plot the princomp object using screeplot() and you can get the
percentage of explained variance from the princomp object p using
p$sdev^2 / sum(p$sdev^2)
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
Peter Dalgaard wrote:
Gad Abraham wrote:
Prof Brian Ripley wrote:
A patch to do this was posted on 2007-09-29 by Glenn Davis. Some
people not addicted to Matlab find the behaviour very inconvenient
and prefer the getline/readline behaviour (triggered by ^R/^S) of
Rterm and R on Unixen
and the input buffer retains the text from
the search, so pressing enter makes the historical command run again.
Try this:
> ls()
character(0)
[ctrl-r]
(reverse-i-search)`l': ls()
[ctrl-c, then enter]
> ls()
character(0)
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Park
Gad Abraham wrote:
If your legal department is still concerned, have them look at the
license in Java. It most likely is GPL (unless you have a special
commercial version of Java installed) and thus makes any Java program
you create subject to the GPL in Java due to the JIT compiler used in
n. We made the
decision because there were other tools comparable to Bison which
already permitted use for non-free programs.''
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.csse.unimelb.e
n sample (from the same population).
Again, see Frank's book for more information about penalisation.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.csse.unimelb.edu.au/~gabraham
Prof Brian Ripley wrote:
This is not to do with 'wide' matrices' but to do with the lack of
column names.
There is a bug -- the definition of 'p' should have NROW and not NCOL.
On Sun, 18 May 2008, Gad Abraham wrote:
I've submitted a bug report:
predict.
#x27; does not have the correct number of columns
> dim(x2)
[1] 5 20
> dim(p$rotation)
[1] 20 5
All predict.prcomp really does is x2 %*% p$rotation, which is a
perfectly legal operation that should yield a 5x5 matrix:
> dim(x2 %*% p$rotation)
[1] 5 5
Thanks,
Gad
--
Gad Abraha
nsole/terminal.
One thing to note: a dataset of 2GB on disk may take much more than 2GB
of RAM when loaded into R, due to the overhead of the metadata and the
fact that pointers are 64-bit long as well.
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Vi
s2$u,1,1) %*% diag(c(s1$d,0,0)) %*% t(s2$v)
all.equal(x,x1,x2)
Yep, you're absolutely correct. As someone's explained to me, the number
of PCs cannot be larger than the rank of the matrix, therefore like you
say the rest of the eigenvalues are practically zero anyway.
Thanks,
Gad
rows? Is the
eigen-decomposition of the covariance matrix really that numerically
bad? (i.e., eigen(cov(x)) )
Thanks,
Gad
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http:/
Charles C. Berry wrote:
> On Tue, 22 Apr 2008, Gad Abraham wrote:
>
>> Charles C. Berry wrote:
>>> On Sat, 19 Apr 2008, Gad Abraham wrote:
>>>
>>>> Charles C. Berry wrote:
>>>>> On Fri, 18 Apr 2008, Gad Abraham wrote:
>>&
Charles C. Berry wrote:
> On Sat, 19 Apr 2008, Gad Abraham wrote:
>
>> Charles C. Berry wrote:
>>> On Fri, 18 Apr 2008, Gad Abraham wrote:
>>>
>>> > Frank E Harrell Jr wrote:
>>> > > Gad Abraham wrote:
>>> > > &g
Prof Brian Ripley wrote:
> On Sun, 20 Apr 2008, Gad Abraham wrote:
>
>> Hi,
>>
>> Say x.train is a matrix of covariates that I want to do PCA on, so I can
>> do regression on its principal components, and x.test is a test set of
>> the same covariates on whi
es:
x.test.pc <- predict(p, newdata=x.test, center=TRUE, scale=TRUE)
Thanks,
Gad
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.csse.unimelb.edu.au/~gabraham
__
un R.
RPy justs translates between Python and R, it's not an implementation of
R in Python, so you have to have R installed somewhere for it to work.
What is possible is to use a web-services approach, where the GUI sends
requests to a web server that has Python, RPy, and R installed on it,
gets
Charles C. Berry wrote:
> On Fri, 18 Apr 2008, Gad Abraham wrote:
>
>> Frank E Harrell Jr wrote:
>>> Gad Abraham wrote:
>>>> Hi,
>>>>
>>>> Design isn't strictly an R base package, but maybe someone can explain
>>>> the fo
Frank E Harrell Jr wrote:
> Gad Abraham wrote:
>> Hi,
>>
>> Design isn't strictly an R base package, but maybe someone can explain
>> the following.
>>
>> When lrm is called within a function, it can't find the dataset dd:
>>
>> >
base packages:
[1] splines stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] Design_2.1-1 survival_2.34 Hmisc_3.4-3
loaded via a namespace (and not attached):
[1] cluster_1.11.9 grid_2.6.2 lattice_0.17-4 rcompgen_0.1-17
--
Gad Abraham
quot; operator instead, see a post by Bill Venables a few
days ago called ``What to use for assignment, " = " or " <- "?'')
whereas "==" is the binary operator for testing equality, as in
> x == 3
[1] TRUE
Have a look at any introductory R book or at
asy way of
extracting the number of observations per cluster from hclust (except by
assigning the new levels to the data and then counting the occurrences).
Does anyone know of code that does this already?
Thanks,
Gad
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010
does not detect that, whereas lm() would. I'd suggest ols()
> is to blame here.
Thanks for clarifying that.
Gad
--
Gad Abraham
Dept. CSSE and NICTA
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web:
LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_AU.UTF-8;LC_IDENTIFICATION=C
attached base packages:
[1] splines stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] Design_2.1-1 survival_2.34 Hmisc_3.4-3
loaded via a namespace (and not attached):
[
> Gad Abraham wrote:
>> Hi,
>> In the boot package, the original statistic is simply the statistic
function evaluated on the original data (called t0).
>> However, in Harrell et al 1996 "Multivariable prognostic models..."
Stats Med vol 15, pp. 361--387, it is diffe
nction to
the bootstrap sample. Then optimism O = average(D_boot - D_orig),
and the bias-corrected estimate is D_app - O.
Can someone explain this difference, and especially why boot
doesn't evaluate the frozen model again on the original data?
Thanks,
Gad
--
Gad Abraham
Departmen
Duncan Murdoch wrote:
> Gad Abraham wrote:
>> Hi,
>>
>> In 2.6.0 arima() used to return an object with attribute `x' which is
>> the observed time series, but this has been dropped from 2.6.1.
>>
>> This breaks the forecast function in package for
hanks,
Gad
--
Gad Abraham
Department of Mathematics and Statistics
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.ms.unimelb.edu.au/~gabraham
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed t
Charles C. Berry wrote:
> On Sat, 8 Dec 2007, Charles C. Berry wrote:
>
>> Gad Abraham ms.unimelb.edu.au> writes:
>>
>>> Hi,
>>>
>>> I'm using natural cubic splines from splines::ns() in survival
>>> regression (regressing inter-arri
y bad here because the predictions grow without bound.
Setting the boundary knots to extreme values such as (0, 1) doesn't
help, because there are no intermediate knots to constrain the spline
and you get a very large cubic curve between these boundaries and the
observed sizes.)
Any suggest
Charles C. Berry wrote:
> On Wed, 5 Dec 2007, Gad Abraham wrote:
>
>> Hi,
>>
>> The following error looks like a bug to me but perhaps someone can shed
>> light on it:
>>
>> > library(splines)
>> > library(survival)
>> > s <
e (and not attached):
[1] rcompgen_0.1-17
--
Gad Abraham
Department of Mathematics and Statistics
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.ms.unimelb.edu.au/~gabraham
__
R-help@r-projec
west AIC,
> or by the lowest standard deviation of the residuals ?
I've found that for a given time series you can fit several different
ARIMA models with very similar results. Out of a group of "sensible"
models (judged by residuals and cross-validated forecast MSE), I'd
David wrote:
> Hi
> Which package(s) is R fit frailty models to univariate survival data, i.e.
> simple data with one survival time per person.
I haven't used it myself, but have a look at ?frailty in the survival
package.
--
Gad Abraham
Department of Mathematics and Statistics
interest rate on time
and indicator variables for day of week and special days like holidays.
Then fit an ARIMA to the regression residuals.
* Any specific reason why ARMA(1,1)? Have you looked at the acf and pacf
of the time series?
Cheers,
Gad
--
Gad Abraham
Department of Mathematics and St
a look at
http://www.adrem.ua.ac.be/~goethals/software/
Cheers,
Gad
--
Gad Abraham
Department of Mathematics and Statistics
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.ms.unimelb.edu.au/~gabraham
Thomas Lumley wrote:
> On Tue, 16 Oct 2007, Gad Abraham wrote:
>
>> Gad Abraham wrote:
>>> Hi,
>>>
>>> I'm using survreg() from the survival package for parametric survival
>>> regression (modelling inter-arrival times of patients to a waitin
Terry Therneau wrote:
> Gad Abraham wrote:
>> Hi,
>>
>> I'm using survreg() from the survival package for parametric survival
>> regression (modelling inter-arrival times of patients to a waiting list
>> as exponentially distributed, with various regres
Gad Abraham wrote:
> Hi,
>
> I'm using survreg() from the survival package for parametric survival
> regression (modelling inter-arrival times of patients to a waiting list
> as exponentially distributed, with various regressors such as queue size
> and season).
>
this?
Thanks,
Gad
--
Gad Abraham
Department of Mathematics and Statistics
The University of Melbourne
Parkville 3010, Victoria, Australia
email: [EMAIL PROTECTED]
web: http://www.ms.unimelb.edu.au/~gabraham
__
R-help@r-project.org mailing list
73 matches
Mail list logo