I know someone who has R code for SF-36 and perhaps SF-12. Aren't there
copyright issues relating to SF-* even if it is reprogrammed?
Frank
-
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context:
http://r.789695.n4.nabble.com/SF-36
Yes the company behind that probably received federal funds for some of the
research and has been very careful to minimize their contribution to the
community.
I didn't understand your parenthetical remark.
Frank
-
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View
CRAN has a significant update to rms. Windows and unix/linux versions
are available and I expect the Mac version to be available soon.
The most significant improvement is addition of latex=TRUE
arguments to model fitting print methods, made especially for use
with Sweave.
Here is a summary
. Nabble sent a link to
turn off e-mail but the r-help mail service rejected nabble's command.
Thanks
Frank
-
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context:
http://r.789695.n4.nabble.com/Where-to-find-R-help-options-web-page-tp2535123p2535123.html
(and to be
able to post on it).
Frank
-
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context:
http://r.789695.n4.nabble.com/Where-to-find-R-help-options-web-page-tp2535123p2535148.html
Sent from the R help mailing list archive at Nabble.com
Aha. Thanks David. I somehow thought that that whole section was for
administrators only.
Frank
-
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context:
http://r.789695.n4.nabble.com/Where-to-find-R-help-options-web-page-tp2535123p2535172.html
Baseline should appear only as a baseline and should be removed from
the set of longitudinal responses. This is often done with a merge( )
operation.
Frank
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt
the time to consult a statistician is before you have done any
statistical analysis.
Frank Harrell
I have a couple of Kleinbaum's (et al) other texts and find them to be
well written and reasoned, so I suspect the citation above would be as
accessible as any.
Thank you, that is useful
The absence of stepwise methods works to your advantage, as these
yield invalid statistical inference and inflated regression
coefficients.
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Thu,
On Thu, 2 Sep 2010, stephenb wrote:
sorry to bump in late, but I am doing similar things now and was browsing.
IMHO anova is not appropriate here. it applies when the richer model has p
more variables than the simpler model. this is not the case here. the
competing models use different
don't feel bad. You just know that something somewhere is
probably wrong with the model. I focus on directed tests such as
allowing all continuous variables to have nonlinear effects or
allowing selected interactions, and finding out how important the
complex model terms are.
Frank Harrell
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Mon, 30 Aug 2010, Ravi Varadhan wrote:
Hi,
I fit a Cox PH model to estimate the cause-specific hazards (in a competing
risks setting). Then , I
Update to the rms package which is the version now being actively
supported. New features will not be added to Design. The nomogram
function in rms separates the plotting into a plot method for easier
understanding. You can control all axes - some experimentation can
tell you if you can
Samuel,
Since the difference in AUCs has insufficient power and doesn't really
take into account the pairing of predictions, I recommend the Hmisc
package's rcorrp.cens function. Its method has good power and asks
the question is one predictor more concordant than the other in the
same
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Sat, 21 Aug 2010, Donald Paul Winston wrote:
Sweave and LaTex is way to much overhead to deal with. There should be a
built in standard report()
Your notes are not well thought out.
You'll find that r-help is a friendly place for new users that do not
come in with an attitude.
I once used SAS (for 23 years) and know it very well. I wrote the
first SAS procedures for a graphics device, percentiles,
logistic regression, and Cox
, Frank Harrell wrote:
Your notes are not well thought out.
You'll find that r-help is a friendly place for new users that do not come in
with an attitude.
I once used SAS (for 23 years) and know it very well. I wrote the first SAS
procedures for a graphics device, percentiles, logistic
On Sat, 21 Aug 2010, Donald Paul Winston wrote:
Good grief. Adding a report function is not going to make R less flexible. Don't
you want to use a tool that's relevant to the rest of the world? That world is
much bigger then your world. This is ridiculous.
Looks like some people are
It would be good to tell us of the frequency of observations in each
category of Y, and the number of continuous X's. Recursive
partitioning will require perhaps 50,000 observations in the less
frequent Y category for its structure and predicted values to
validate, depending on X and the
On Fri, 20 Aug 2010, Kay Cichini wrote:
hello,
my data-collection is not yet finished, but i though have started
investigating possible analysis methods.
below i give a very close simulation of my future data-set, however there
might be more nominal explanatory variables - there will be no
What do low level proc print and proc report have on Sweave or
http://biostat.mc.vanderbilt.edu/wiki/pub/Main/StatReport/summary.pdf?
If proc print and proc report are 4G, let's move back a generation.
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
At the heart of this you have a problem in incomplete conditioning.
You are computing things like Prob(X x) when you know X=x. Working
with a statistician who is well versed in probability models will
undoubtedly help.
Frank
Frank E Harrell Jr Professor and ChairmanSchool of
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Tue, 17 Aug 2010, Rob James wrote:
1) How does one capture the plots from the plsmo procedure? Simply
inserting a routing call to a graphical
The values of slentry and slstay that will avoid ruining the
statistical properties of the result are slentry=1.0 and slstay=1.0.
Frank
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Sat, 14 Aug
Once you guys figure all this out, I'm glad to modify bplot to pass
more arguments lattice if needed.
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Fri, 13 Aug 2010, David Winsemius wrote:
install.packages('rms')
require(rms)
?Gls
?plot.Predict
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Sat, 14 Aug 2010, Camilo Mora wrote:
Hi everyone:
Is there a function in R to calculate
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Sat, 14 Aug 2010, David Winsemius wrote:
On Aug 14, 2010, at 9:59 AM, Frank Harrell wrote:
Once you guys figure all this out, I'm glad
David,
In the Cox and many other regression models, the effect of a variable
is context-dependent. There is an identifiability problem in what you
are doing, as discussed by
@ARTICLE{for95mod,
author = {Ford, Ian and Norrie, John and Ahmadi, Susan},
year = 1995,
title = {Model
Please check the code. I hope that Brier is on the uncalibrated
probabilities.
Calibrated probabilities are from 1/(1+exp(-[a+b logit(uncalibrated
probs)]) where a and b are maximum likelihood estimators (they will be
0 and 1 in training data).
Frank
Frank E Harrell Jr Professor and
.
_
De : Bert Gunter gunter.ber...@gene.com
À : Biau David djmb...@yahoo.fr
Cc : Frank Harrell f.harr...@vanderbilt.edu; r help list
r-help@r-project.org
Envoyé le : Ven 13 août 2010, 18h 22min 58s
Objet : Re: [R] Re : How to compare the effect of a variable across regression
models
On Wed, 11 Aug 2010, david dav wrote:
Hi,
I would like to extract the coefficients of a logistic regression
(estimates and standard error as well) in lrm as in glm with
summary(fit.glm)$coef
Thanks
David
coef(fit)
sqrt(diag(vcov(fit)))
But these will not be very helpful except in the
tests to be exactly the same?
Mark - you can see the code for this at the bottom of anova.rms.
Compute W, divide by numerator d.f., then compute P-value using F with
numerator and error d.f.
Frank
Thanks in advance; I really appreciate any help you can give.
Mark
Frank Harrell wrote
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Wed, 11 Aug 2010, Michal Figurski wrote:
Peter, Frank, David and others,
Thank you all for your ideas. I understand your lack of trust in PB
University
On Wed, 11 Aug 2010, S Ellison wrote:
Frank Harrell f.harr...@vanderbilt.edu 11/08/2010 17:02:03
This problem seems to cry out for one of the many available robust
regression methods in R.
Not sure that would be much more appropriate, although it would
_appear_ to work. The PB
Classification accuracy is an improper scoring rule, and one of the
problems with it is that the proportional classified correctly can be
quite good even if the model uses no predictors. [Hence omitting the
intercept is also potentially problematic.]
Frank E Harrell Jr Professor and
Please give the prescription. The article is not available on our
extensive online library. I wonder if the method can compete with the
bootstrap.
Frank
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt
Research Laboratory
3400 Spruce St. 7 Maloney
Philadelphia, PA 19104
tel. (215) 662-3413
On 2010-08-10 12:29, Frank Harrell wrote:
Please give the prescription. The article is not available on our
extensive online library. I wonder if the method can compete with the
bootstrap.
Frank
Frank E
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Mon, 9 Aug 2010, Harsh wrote:
Hello useRs,
I have a problem at hand which I'd think is fairly common amongst
groups were R is being adopted for
Note that stepwise variale selection based on AIC has all the problems
of stepwise variable selection based on P-values. AIC is just a
restatement of the P-Value.
Frank
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics
Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Mon, 9 Aug 2010, Kingsford Jones wrote:
On Mon, Aug 9, 2010 at 10:27 AM, Frank Harrell f.harr...@vanderbilt.edu wrote:
Note that stepwise variale selection based on AIC has all
On Mon, 9 Aug 2010, Alexander Eggel wrote:
Hello everybody,
I need to know which samples (S1-S6) contain a value that is bigger than the
median + five standard deviations of the column he is in. This is just an
Why not the 70th percentile plus 6 times the difference in the 85th
and 75th
On Mon, 9 Aug 2010, Mark Seeto wrote:
Hello, I have a general question about combining imputations as well as a
question specific to the rms and Hmisc packages.
The situation is multiple regression on a data set where multiple
imputation has been used to give M imputed data sets. I know how
There are many ways to do this. Here is one.
install.packages('rms')
require(rms)
dd - datadist(x, y); options(datadist='dd')
f - ols(z ~ x + y)
plot(Predict(f))# plot all partial effects
plot(Predict(f, x)) # plot only the effect of x
plot(Predict(f, y)) # plot only the effect of y
f -
In an upcoming release of the rms package, all fit objects can be
printed using LaTeX if putting LaTeX code directly to the console
(this is optimized for Sweave). You will be able to say print(fit,
latex=TRUE).
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
To add to David's note, the Kruskal-Wallis test is the nonparametric
counterpart to one-way ANOVA. You can get a series of K-W tests for
several grouping or continuous independent variables (but note these
are SEPARATE analyses) using the Hmisc package's spearman2 function.
The generalization
In addition the poster did not tell us what is wrong with a
nonparametric test.
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Mon, 2 Aug 2010, Bert Gunter wrote:
My sympathies, but I don't
Hi Rob,
rms wants symmetry in the sense that the interactions need to use the
same number and location of spline knots as the main effects. So if
using the * notation omit the main effects (which are generated
automatically) and live with the equal knots. Or use the restricted
interaction
, method=bands)
That looks OK but I can't test it right now. Please continue to have
a look, and if you still don't see the problem provide a tiny
reproducible example with self-contained data I can access.
Frank
Thanks for your help.
KM
On Jul 27, 9:58 pm, Frank Harrell f.harr
This is true by definition.
Read about the bootstrap which may give you some good background
information.
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
On Wed, 28 Jul 2010, xin wei wrote:
hi,
If the x-axis variable is really a factor, xYplot will not handle it.
You probably need a dot chart instead (see Hmisc's Dotplot).
Note that it is unlikely that the confidence intervals are really
symmetric.
Frank
On Tue, 27 Jul 2010, Kang Min wrote:
Hi,
I'm trying to plot a graph with
Easiest thing is to sample with replacement from the original data.
This is the idea behind the bootstrap, which is sampling from the
empirical CDF.
Frank E Harrell Jr Professor and ChairmanSchool of Medicine
Department of Biostatistics Vanderbilt University
This has been shown to yield unreliable analyses. Use the more formal
proportional odds ordinal logistic model. This is a generalization of
the Wiloxon-Mann-Whitney-Kruskal-Wallis statistic. This is
implemented in the rms package and elsewhere.
Frank E Harrell Jr Professor and Chairman
401 - 452 of 452 matches
Mail list logo
