Re: [R] Hmisc describe error
The error is just a misleading error message. loading the data produced column sdate as str(prostate) $ sdate :Classes 'labelled', 'dates' atomic [1:502] 2778 2820 2933 2999 3002 ... .. ..- attr(*, format)= symbol ddmmmyy .. ..- attr(*, label)= chr Date on study prostate$sdate - as.Date(prostate$sdate,origin=1970-01-01) # fixes the problem and 'describe' runs just as a comment: maybe the R dataset should have sdate as a Date class. -- View this message in context: http://r.789695.n4.nabble.com/Hmisc-describe-error-tp4644708p4644711.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
[R] odfWeave fails to load
R version 2.15.0 (2012-03-30) Copyright (C) 2012 The R Foundation for Statistical Computing ISBN 3-900051-07-0 Platform: i386-pc-mingw32/i386 (32-bit) package ‘survey’ successfully unpacked and MD5 sums checked package ‘odfWeave.survey’ successfully unpacked and MD5 sums checked library(odfWeave.survey) Loading required package: odfWeave Error: package ‘odfWeave’ could not be loaded any ideas, anybody?? I had odfSweave on 2.12, but no such thing on 2.15 just odfWeave.survey and it won't load. -- View this message in context: http://r.789695.n4.nabble.com/odfWeave-fails-to-load-tp4631700.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] aggregate with cumsum
Bill, what will be the fastest way to output not just single lines but small data frames of about 60 rows? I prefer writing to a text file because the final output is large 47k times 60 rows and since I do not know the size of it I have to use rbind to build the object which creates the memory problems described here: http://www.matthewckeller.com/html/memory.html look at the swiss cheese paragraph. kind regards Stephen -- View this message in context: http://r.789695.n4.nabble.com/aggregate-with-cumsum-tp2992383p3324610.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] Prediction confidence intervals for a Poisson GLM
I am following up on an old post. Please, comment: it appears that predict(glm.model,type=response,se.fit=T) will do all the conversions and give se on the scale of the response. This only takes into account the error in parameter estimation. what a prediction interval is meant to be usually means it has to capture the error due to both parameter estimation and sampling variation ie it encompasses the actual realizations. for a given parameter there is sampling variation and that is not included in the output of predict. the discreteness of models makes it quite difficult to estimate a percentile interval, though. for binary outcomes, I think it does not make sense. for Poisson and binomial (grouped binary) I think it is possible to get approximations at least and this is what the original poster needed I think. so, let's say we have plow and pup for an observation from predict. if size=100 for that obs. predlow=qbinom(.025,100,plow) predup=qbinom(.975,100,pup) will give the prediction bounds. this I think partly ignores possible overdispersion. Please, suggest a better way taking overdispersion into account (in the qbinom part). Thanks everybody. Stephen B. -- View this message in context: http://r.789695.n4.nabble.com/Prediction-confidence-intervals-for-a-Poisson-GLM-tp841577p2533070.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] Comparing COXPH models, one with age as a continuous variable, one with age as a three-level factor
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 variables. you are left with IC. by transforming a continuous variable into categorical you are smoothing, which is the idea of GAM. if you look at what is offered in GAMs you may find better approximations f(age) as well as tools for testing among different f(age) transformations. regards. S. -- View this message in context: http://r.789695.n4.nabble.com/how-to-get-design-matrix-tp891987p2524289.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] Data Manipulations and SQL
Greetings Gabor, is it possible to open a channel to a data frame in the default environment? there are cases when using a sql update statement is the simplest alternative, so instead of dumping the df and then updating and then reimporting it I would like to update the df directly in R. Thank you. Stephen -- View this message in context: http://r.789695.n4.nabble.com/Data-Manipulations-and-SQL-tp860420p2340098.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] unable to get bigglm working, ATTN: Thomas Lumley
the model fails to converge after more than 3 hours ( I went home so don't know how long it took) bigglm (formula = resp ~ relage+I(relage^2)+termfac+ri , + data = a, family = binomial(link='logit')); Large data regression model: bigglm(formula = resp ~ relage + I(relage^2) + termfac + ri, data = a, family = binomial(link = logit)) Sample size = 12758187 failed to converge after 8 iterations Warning message: In bigglm.function(formula = resp ~ relage + I(relage^2) + termfac + : ran out of iterations and failed to converge SAS converges NOTE: PROC LOGISTIC is modeling the probability that resp='1'. NOTE: Convergence criterion (GCONV=1E-8) satisfied. NOTE: There were 12758187 observations read from the data set SRVRUSER.COMMIT. NOTE: The data set WORK.OUT3 has 11 observations and 15 variables. NOTE: PROCEDURE LOGISTIC used (Total process time): real time 2:25.42 cpu time1:16.79 I did not see a trace argument in bigglm. is there another way to see what is happening? Thank you. -- View this message in context: http://r.789695.n4.nabble.com/unable-to-get-bigglm-working-ATTN-Thomas-Lumley-tp2276524p2278381.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] unable to get bigglm working, ATTN: Thomas Lumley
I decided to give it 1 more variable, which is strongly significant to help the optimization and it throws: bigglm (formula = resp ~ relage+relage2+termfac+ri+sn , + data = a, family = binomial(link='logit')); Error in bigglm.function(formula = resp ~ relage + relage2 + termfac + : model matrices incompatible -- View this message in context: http://r.789695.n4.nabble.com/unable-to-get-bigglm-working-ATTN-Thomas-Lumley-tp2276524p2278734.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
[R] unable to get bigglm working, ATTN: Thomas Lumley
I am using an example posted in this help forum to work with a file. the head
of the file looks like:
97 2007-03-05 2007-06-01 90 3 5.450 205500.00 999.00 999.000 0.000 0 0
97 2007-03-06 2007-06-01 90 3 5.450 205500.00 999.00 999.000 0.000 1 0
97 2007-03-07 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 2 0
97 2007-03-08 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 3 0
97 2007-03-09 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 4 0
97 2007-03-12 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 7 0
97 2007-03-13 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 8 0
97 2007-03-14 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 9 0
97 2007-03-15 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 10 0
97 2007-03-16 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 11 0
the code is:
make.data - function (filename, chunksize, ...) {
conn-NULL;
function (reset=FALSE) {
if (reset) {
if (!is.null(conn)) {
close(conn);
};
conn - file (description=filename, open=r);
} else {
rval - read.table (conn, nrows=chunksize,sep=' ',
skip=0, header=FALSE,...);
if (nrow(rval)==0) {
close(conn);
conn-NULL;
rval-NULL;
} else {
rval$relage - rval$loctime/rval$term;
};
return(rval);
}
}
};
a - make.data ( filename = G:/sqldata/newf4.csv, chunksize = 10,
colClasses = list (NULL, Date,Date, integer,
factor,rep(numeric,5),rep(integer,2)),
col.names = c(id,dt, promdt,term, termfac,
commintr,commbal,issuebal,intr,ri,loctime,resp)
)
library(biglm);
bigglm (formula = resp ~ poly(relage,2,raw=TRUE)+termfac+ri ,
data = a, family = binomial(link='logit'));
### output:
bigglm (formula = resp ~ poly(relage,2,raw=TRUE)+termfac+ri ,
+ data = a, family = binomial(link='logit'));
Error in is(object, Class) :
trying to get slot className from an object of a basic class (list)
with no slots
### the following can create a df, so the problem is not loading the data
(maybe :-)
a - read.table ( G:/sqldata/newf4.csv, nrows= 50, sep=' ',head=F,
colClasses = c(NULL,
Date,Date,integer,factor,rep(numeric,5),rep(integer,2)),
col.names = (id,dt, promdt,term, termfac,
commintr,commbal,issuebal,intr,ri,loctime,resp)
)
Thanks everybody.
--
View this message in context:
http://r.789695.n4.nabble.com/unable-to-get-bigglm-working-ATTN-Thomas-Lumley-tp2276524p2276524.html
Sent from the R help mailing list archive at Nabble.com.
__
R-help@r-project.org 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.
Re: [R] unable to get bigglm working, ATTN: Thomas Lumley
Sincere apologies and many thanks. The pasted code from the help forum has the
error and I did not see it. When I wrote my own import I used c as I should.
See:
http://r.789695.n4.nabble.com/R-Example-function-for-bigglm-biglm-data-input-from-file-td816496.html#a816496
Stephen Bond
| Senior Analyst | Treasury Analytics | 416-956-3092
From: Thomas Lumley [via R]
[mailto:ml-node+2276661-800075919-75...@n4.nabble.com]
Sent: Friday, July 02, 2010 12:38 PM
To: Bond, Stephen
Subject: Re: unable to get bigglm working, ATTN: Thomas Lumley
Actually, I think the problem *is* reading in the data
If I try reading in your supplied lines of data with the read.table arguments()
in your make.data() function I get your error message.
data-read.table(tmp-textConnection(
+ 97 2007-03-05 2007-06-01 90 3 5.450 205500.00 999.00 999.000 0.000 0 0
+ 97 2007-03-06 2007-06-01 90 3 5.450 205500.00 999.00 999.000 0.000 1 0
+ 97 2007-03-07 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 2 0
+ 97 2007-03-08 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 3 0
+ 97 2007-03-09 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 4 0
+ 97 2007-03-12 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 7 0
+ 97 2007-03-13 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 8 0
+ 97 2007-03-14 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 9 0
+ 97 2007-03-15 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 10 0
+ 97 2007-03-16 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 11 0
+ ), colClasses = list (NULL, Date,Date, integer,
+ factor,rep(numeric,5),rep(integer,2)),
+ col.names = c(id,dt, promdt,term, termfac,
+ commintr,commbal,issuebal,intr,ri,loctime,resp)
+ )
Error in is(object, Class) :
trying to get slot className from an object of a basic class (list) with
no slots
My guess is that the problem is that you use list() instead of c() for
constructing your colClasses argument. In the code for reading the file that
you didn't have a problem with, you used c().
-thomas
On Fri, 2 Jul 2010, stephenb wrote:
I am using an example posted in this help forum to work with a file. the head
of the file looks like:
97 2007-03-05 2007-06-01 90 3 5.450 205500.00 999.00 999.000 0.000 0 0
97 2007-03-06 2007-06-01 90 3 5.450 205500.00 999.00 999.000 0.000 1 0
97 2007-03-07 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 2 0
97 2007-03-08 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 3 0
97 2007-03-09 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 4 0
97 2007-03-12 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 7 0
97 2007-03-13 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 8 0
97 2007-03-14 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 9 0
97 2007-03-15 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 10 0
97 2007-03-16 2007-06-01 90 3 5.450 205500.00 999.00 999.000 -0.100 11 0
the code is:
make.data - function (filename, chunksize, ...) {
conn-NULL;
function (reset=FALSE) {
if (reset) {
if (!is.null(conn)) {
close(conn);
};
conn - file (description=filename, open=r);
} else {
rval - read.table (conn, nrows=chunksize,sep=' ',
skip=0, header=FALSE,...);
if (nrow(rval)==0) {
close(conn);
conn-NULL;
rval-NULL;
} else {
rval$relage - rval$loctime/rval$term;
};
return(rval);
}
}
};
a - make.data ( filename = G:/sqldata/newf4.csv, chunksize = 10,
colClasses = list (NULL, Date,Date, integer,
factor,rep(numeric,5),rep(integer,2)),
col.names = c(id,dt, promdt,term, termfac,
commintr,commbal,issuebal,intr,ri,loctime,resp)
)
library(biglm);
bigglm (formula = resp ~ poly(relage,2,raw=TRUE)+termfac+ri ,
data = a, family = binomial(link='logit'));
### output:
bigglm (formula = resp ~ poly(relage,2,raw=TRUE)+termfac+ri ,
+ data = a, family = binomial(link='logit'));
Error in is(object, Class) :
trying to get slot className from an object of a basic class (list)
with no slots
### the following can create a df, so the problem is not loading the data
(maybe :-)
a - read.table ( G:/sqldata/newf4.csv, nrows= 50, sep=' ',head=F,
colClasses = c(NULL,
Date,Date,integer,factor,rep(numeric,5),rep(integer,2)),
col.names = (id,dt, promdt,term, termfac,
commintr,commbal,issuebal,intr,ri,loctime,resp)
)
Thanks everybody.
--
View this message in context:
http://r.789695.n4.nabble.com/unable-to-get-bigglm-working-ATTN-Thomas-Lumley-tp2276524p2276524.htmlhttp://r.789695.n4.nabble.com/unable-to-get-bigglm-working-ATTN-Thomas-Lumley-tp2276524p2276524.html?by-user=t
Sent from the R help mailing list archive at Nabble.com.
__
[hidden email]/user/SendEmail.jtp?type=nodenode=2276661i=0 mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
Re: [R] unable to get bigglm working, ATTN: Thomas Lumley
after correcting the error spotted by Thomas, I tried again and it goes to work, but there is no result after 1 hour. is there anything I can do to debug? the dataset has 12mln rows and SAS on a server will produce results in 15 secs. I am running this on PC with XP (hence mem limit of 2.5 gigs) fresh after restart and no other applications. thank you. Stephen -- View this message in context: http://r.789695.n4.nabble.com/unable-to-get-bigglm-working-ATTN-Thomas-Lumley-tp2276524p2276747.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] Example function for bigglm (biglm) data input from file
Richard, do you have an example for an ODBC connection? Thank you Stephen -- View this message in context: http://r.789695.n4.nabble.com/R-Example-function-for-bigglm-biglm-data-input-from-file-tp816496p2230710.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
[R] multiplicative contraint in LS
I have a model that can be estimated in nls
mod.base -
deriv(~(Jan+a1*Feb+a2*Mar+a3*Apr+a4*May+a5*Jun+a6*Jul+a7*Aug+a8*Sep+a9*Oct+a10*Nov+a11*Dec)*(b2*relAge+b3*relAge^2),
c(a1,a2,a3,a4,a5,a6,a7,a8,a9,a10,a11,b2,b3),
function(a1,a2,a3,a4,a5,a6,a7,a8,a9,a10,a11,b2,b3,relAge,Jan,Feb,Mar,Apr,May,Jun,Jul,Aug,Sep,Oct,Nov,Dec){})
nls1 -
nls(y~mod.base(a1,a2,a3,a4,a5,a6,a7,a8,a9,a10,a11,b2,b3,relAge,Jan,Feb,Mar,Apr,May,Jun,Jul,Aug,Sep,Oct,Nov,Dec),data=fix1,
start=list(a1=.99,a2=1.01,a3=1.02,a4=1.04,a5=1,a6=1.05,a7=1.06,a8=1.07,a9=1.08,a10=1.09,a11=1.1,b2=.02,b3=-.02))
I would like to estimate with lm/gls to use the greater
decoration/diagnostics available in lm/gls
calling
lm(y~poly(relAge,2,raw=T)*sn-sn-poly(relAge,2,raw=T)-1,data=fix1)
gives me parameters that are not constrained across season eg when I divide
poly(relAge, 2, raw = T)1:snFeb / poly(relAge, 2, raw = T)1:snJan
I get a number which is not identical to
poly(relAge, 2, raw = T)2:snFeb / poly(relAge, 2, raw = T)2:snJan
which is to be expecetd as I do not constrain. Please, help with
constraining.
The obvious reparametrization leads me to the nls, which does not have
correlation and variance funcs as gls.
Thank you
--
View this message in context:
http://n4.nabble.com/multiplicative-contraint-in-LS-tp1589181p1589181.html
Sent from the R help mailing list archive at Nabble.com.
__
R-help@r-project.org 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.
Re: [R] Problem with rcmd SHLIB
I have never got it working with rcmd SHLIB. it is much easier to create a shared library separately with the C compiler and then load it with dyn.load the makefiles for SHLIB are quite big and going through them will give you a headache, so just use the flag for a dll when you compile. -- View this message in context: http://n4.nabble.com/Problem-with-rcmd-SHLIB-tp1589180p1589188.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] RPostgreSQL: unable to load shared library
I am trying the same as the post above and the actual error is:
the ordinal 284 could not be located in the dynamic link library
SSLEAY32.dll
I searched for that dll and it is not to be found. windows forums provide
vague directions to reinstall Windows, which I cannot do because this is a
company PC and IT will be happy to tell me that's why open source programs
are not supported and not encouraged.
any suggestions are appreciated.
Stephen
Prof Brian Ripley wrote:
On Wed, 14 Oct 2009, Fanfaar wrote:
Hello list,
I'm using R 2.9.2 on a WinXP system, and I installed the RPostgreSQL
library using the package installer. When trying to load it, I get the
following error:
library('RPostgreSQL')
Error in inDL(x, as.logical(local), as.logical(now), ...) :
unable to load shared library
'C:/PROGRA~1/R/R/library/RPostgreSQL/libs/RPostgreSQL.dll':
LoadLibrary failure: The operating system cannot run %1.
Error: package/namespace load failed for 'RPostgreSQL'
So one way or the other, the DLL is not found... Does anyone know how
That is not what it says: it says it cannot *load* the DLL. You need
the PostgreSQL client dll in your path, and I guess that (or its
version) is the problem. (Usually Windows gives you a popup with more
information, and indeed on my laptop it told me LIBPQ.DLL could not be
found.) And pedump suggests that it is linked against entry points by
number not name, a very fragile arrangement.
I always worry that packages that link to external DLLs can be very
dependent on the version of that DLL (and see the above comment). I
could not see a description of the version of PostgreSQL used on Uwe's
ReadMe (assuming this is a binary from CRAN), and suggest (as did the
rw-FAQ) that you install RPostgreSQL from source against your own
PostgreSQL installation. (That's what I do on my Windows desktop
which does have PostgreSQL installed, and when I updated PostgreSQL I
had to re-install PostgreSQL )
to fix this? I don't suppose DLL should be directly in my PATH, right?
Thanks for any hints,
Arnout
__
R-help@r-project.org 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.
Yes another person who thinks that does not apply to them.
--
Brian D. Ripley, rip...@stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UKFax: +44 1865 272595
__
R-help@r-project.org 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.
--
View this message in context:
http://old.nabble.com/RPostgreSQL%3A-unable-to-load-shared-library-tp25895592p26197137.html
Sent from the R help mailing list archive at Nabble.com.
__
R-help@r-project.org 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.
Re: [R] R and Python
Not only RSPython does not support Windows, but it won't compile on most Unix platforms and fixing it may be more time than writing a package on your own. Like the other post, I use Python for pre-processing and then load into R. If you need text processing with line by line access you can also use Java, which will communicate to R. cheers Stephen Ryan Sheftel-2 wrote: I am a long time user of R for financial time series analysis (xts, zoo, etc) and for my next project I am thinking of adding the Python language to the mix. The reason for adding Python is primarily its non-statistical capabilities. So my questions are what have people's experiences been with using interop between R and Python. I see there are two items, rPy and RSPython. It looks like rPy makes it possible to call R code from Python, and RSPython goes both ways. My needs would be to use Python to drive R to get it's extensive time series and stats, and also to get to Python objects from inside R. I searched the list archives and it looks like many people use rPy, but what about RSPython for calling Python from R? It looks like rPy only goes from R into Python, and RSPython has not been updated since 2005? Other messages in the archives state that RSPython only works on Unix? Would I be foolish to build anything mission critical on RSPython? Is there a better way to get at Python from R? Thanks as always. [[alternative HTML version deleted]] __ R-help@r-project.org 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. -- View this message in context: http://old.nabble.com/R-and-Python-tp26166781p26199643.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] glmmPQL model selection
Sorry for the late reply. Just use the first 90% of your data to fit and then predict the last 10% and see which one is better. If the random effects are not good it will become very obvious. If the concern is with fixed effects then just use gls which puts the random effects in the error and select model as usual. Emmanuelle TASTARD wrote: Hi, Im sorry, I know that it is a recurrent question but I have not been able to find the response in the Rhelp archives. I think my data require the use of the glmmPQL function but I do not know how to make the model selection. Since the AIC and log-likelihood are apparently meaningless, how can we select the parameters for a model and compare the models to find which one fits best the data? Thanks a lot Emmanuelle Tastard Emmanuelle TASTARD UMR 5174 'Evolution et Diversité Biologique' Université Paul Sabatier Bat 4R3 31062 TOULOUSE CEDEX 9 France tel : 05 61 55 67 59 [[alternative HTML version deleted]] __ r-h...@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 -- View this message in context: http://www.nabble.com/glmmPQL-model-selection-tp3027224p25151417.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] glmmPQL and variance structure
this is very late, but I saw this now as I am dealing with it now:
I think varPower should not be needed here. The family should be one of the
quasi families eg quasibinomial and that will automatically allow
variance/dispersion to become a function of the fit. This is a feature of
glm inherently, which does not exist in lme, so it is called with varPower
(in lme).
If you see that, pls post whether it works as expected on the original
dataset.
regards
Stephen
Spencer Graves wrote:
Thanks for providing a partially reproducible example. I believe the
error message you cite came from lme. I say this, because I modified
your call to glmmPQL2 to call lme and got the following:
library(nlme)
fit.lme - lme(y ~ trt + I(week 2), random = ~ 1 | ID,
+ data = bacteria, weights=varPower(~1))
Error in unlist(x, recursive, use.names) :
argument not a list
I consulted Pinheiro and Bates (2000) Mixed-Effects Models in S and
S-Plus (Springer, sec. 5.2, p.211) to see that the syntax for varPower
appears to be correct. I removed ~ and it worked, mostly:
fit.lme - lme(y ~ trt + I(week 2), random = ~ 1 | ID,
+ data = bacteria, weights=varPower(1))
Warning message:
- not meaningful for factors in: Ops.factor(y[revOrder], Fitted)
I got an answer, though with a warning and not for the problem you
want to solve. However, I then made this modification to a call to my
own modification of Venables and Ripley's glmmPQL and it worked:
fit. - glmmPQL.(y ~ trt + I(week 2), random = ~ 1 | ID,
+ family = binomial, data = bacteria,
+ weights.lme=varPower(1))
iteration 1
iteration 2
iteration 3
fit.
Linear mixed-effects model fit by maximum likelihood
Data: bacteria
Log-likelihood: -541.0882
Fixed: y ~ trt + I(week 2)
(Intercept) trtdrugtrtdrug+ I(week 2)TRUE
2.7742329 -1.0852566 -0.5896635 -1.2682626
Random effects:
Formula: ~1 | ID
(Intercept) Residual
StdDev: 4.940885e-05 2.519018
Variance function:
Structure: Power of variance covariate
Formula: ~fitted(.)
Parameter estimates:
power
0.3926788
Number of Observations: 220
Number of Groups: 50
This function glmmPQL. adds an argument weights.lme to Venables
and Ripley's glmmPQL and uses that in place of
'quote(varFixed(~invwt))' when provided; see below.
hope this helps.
spencer graves
glmmPQL. -
function (fixed, random, family, data, correlation, weights,
weights.lme, control, niter = 10, verbose = TRUE, ...)
{
if (!require(nlme))
stop(package 'nlme' is essential)
if (is.character(family))
family - get(family)
if (is.function(family))
family - family()
if (is.null(family$family)) {
print(family)
stop('family' not recognized)
}
m - mcall - Call - match.call()
nm - names(m)[-1]
wts.lme - mcall$weights.lme
keep - is.element(nm, c(weights, data, subset, na.action))
for (i in nm[!keep]) m[[i]] - NULL
allvars - if (is.list(random))
allvars - c(all.vars(fixed), names(random),
unlist(lapply(random,
function(x) all.vars(formula(x)
else c(all.vars(fixed), all.vars(random))
Terms - if (missing(data))
terms(fixed)
else terms(fixed, data = data)
off - attr(Terms, offset)
if (length(off - attr(Terms, offset)))
allvars - c(allvars, as.character(attr(Terms, variables))[off
+
1])
m$formula - as.formula(paste(~, paste(allvars, collapse = +)))
environment(m$formula) - environment(fixed)
m$drop.unused.levels - TRUE
m[[1]] - as.name(model.frame)
mf - eval.parent(m)
off - model.offset(mf)
if (is.null(off))
off - 0
w - model.weights(mf)
if (is.null(w))
w - rep(1, nrow(mf))
mf$wts - w
fit0 - glm(formula = fixed, family = family, data = mf,
weights = wts, ...)
w - fit0$prior.weights
eta - fit0$linear.predictor
zz - eta + fit0$residuals - off
wz - fit0$weights
fam - family
nm - names(mcall)[-1]
keep - is.element(nm, c(fixed, random, data, subset,
na.action, control))
for (i in nm[!keep]) mcall[[i]] - NULL
fixed[[2]] - quote(zz)
mcall[[fixed]] - fixed
mcall[[1]] - as.name(lme)
mcall$random - random
mcall$method - ML
if (!missing(correlation))
mcall$correlation - correlation
# weights.lme
{
if(is.null(wts.lme))
mcall$weights - quote(varFixed(~invwt))
else
mcall$weights - wts.lme
}
mf$zz - zz
mf$invwt - 1/wz
mcall$data - mf
for (i in 1:niter) {
if (verbose)
cat(iteration, i, \n)
fit - eval(mcall)
etaold - eta
eta -
Re: [R] how to specify two variance effects in gls
I got the answer: weights=varComb(varFixed(~1/n),varPower(~Age)) Bond, Stephen-2 wrote: Hello everybody, I have a dataset where each row has number of subjects and that gives me natural weights for the variance function. Additionally I see that variance increases with Age, which is a regressor. So using gls I have weights=varFixed(~1/n) but don't know how to include the extra effect of the regressor. Fitted values show a quadratic curve vs. age, not sure if that helps. Thanks everybody. Stephen [[alternative HTML version deleted]] __ R-help@r-project.org 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. -- View this message in context: http://www.nabble.com/how-to-specify-two-variance-effects-in-gls-tp25044356p25061604.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] .RData signature
any ideas? has anybody tried to recover RData files before? Duncan Murdoch-2 wrote: On 7/13/2009 3:38 PM, sten...@go.com wrote: Is there a way to identify the beginning and end of an .RData file? I am trying to restore lost files on opensolaris where the table of contents is overwritten and will be reading the raw disk byte by byte. I looked at several files in hex and the end seems different, while most of them, but not all begin with RDX2.X No, there is nothing special written at the end of a file. Many .RData files are compressed (by internal gzip), so the RDX2 header is hidden, and you'll see bytes 1F 8B instead. (I don't think we write the trailer bytes on the gzip stream, but I'm not sure about that.) Duncan Murdoch __ R-help@r-project.org 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. -- View this message in context: http://www.nabble.com/.RData-signature-tp24468047p24472875.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
[R] text() to label points in ggplot
is there a way to label points in a graph using text(locator(1),text) after ggplot() or qplot() ? qplot(date, psavert, data = economics, geom = line,main=jhdjd)-p p+opts(text(locator(1),),new=T) does not work. -- View this message in context: http://www.nabble.com/text%28%29-to-label-points-in-ggplot-tp23545135p23545135.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
Re: [R] Merging data frames, or one column/vector with a data frame filling out empty rows with NA's
You are exceeding your max memory here, so R will not be able to do that. dump both tables into a db such as mysql and then run the query either from RMySQL or from mysql directly. then output the result and import back in R. that will take care of the merge, but not sure what will happen when you actually try to run some stats on the object. it is very likely the operation will exceed memory again. in the end you may have to write your own code which does not attempt to load everything in memory, it could be either R or a lower level language. if you have SAS it will probably work as it deals with large sets in long format well. depending on what you do R may be able to deal with it after a reshape() to a wide format. joe1985 wrote: Hello I have two data frames, SNP4 and SNP1: head(SNP4) Animal MarkerY 3213 194073197 P1001 0.021088 1295 194073197 P1002 0.021088 915 194073197 P1004 0.021088 2833 194073197 P1005 0.021088 1487 194073197 P1006 0.021088 1885 194073197 P1007 0.021088 head(SNP1) AnimalMarker x 3213 194073197 P1001 2 1295 194073197 P1002 1 915 194073197 P1004 2 2833 194073197 P1005 0 1487 194073197 P1006 2 1885 194073197 P1007 0 I want these two data frames merged by 'Marker', but when i try SNP5 - merge(SNP4, SNP1, by = 'Marker', all = TRUE) Error: cannot allocate vector of size 2.4 Gb In addition: Warning messages: 1: In merge.data.frame(SNP4, SNP1, by = Marker, all = TRUE) : Reached total allocation of 1535Mb: see help(memory.size) 2: In merge.data.frame(SNP4, SNP1, by = Marker, all = TRUE) : Reached total allocation of 1535Mb: see help(memory.size) 3: In merge.data.frame(SNP4, SNP1, by = Marker, all = TRUE) : Reached total allocation of 1535Mb: see help(memory.size) 4: In merge.data.frame(SNP4, SNP1, by = Marker, all = TRUE) : Reached total allocation of 1535Mb: see help(memory.size) And error occurs. What i want is the column SNP1$x merged together with SNP4 by Marker, so some markers will have NA's in the 'x'-column in the SNP5 dataset. I also tried this SNP5 - merge(SNP4, SNP1$x, by.x = 'Marker', by.y = 'Marker', all = TRUE) Error in fix.by(by.y, y) : 'by' must specify valid column(s) I won't work either. Does anyone have any idea how to solve this. Regards, Johannes. -- View this message in context: http://www.nabble.com/Merging-data-frames%2C-or-one-column-vector-with-a-data-frame-filling-out-empty-rows-with-NA%27s-tp23171110p23259062.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
[R] dll problem
I am trying to load a dll compiled with the latest cygwin into R 2.7.0 R version 2.7.0 (2008-04-22) Copyright (C) 2008 The R Foundation for Statistical Computing ISBN 3-900051-07-0 I have: Sys.getenv(Path) Path C:\\Windows\\system32;C:\\Windows;C:\\Windows\\System32\\Wbem;C:\\Program Files\\MBTrading\\MBT Navigator;C:\\Program Files\\R\\R-2.7.0\\bin;C:\\cygwin\\bin the source is: !!! f1.f SUBROUTINE f1(a) INTEGER :: a, i a=2 RETURN END SUBROUTINE f1 compiled by g77 -shared -o f1.dll f1.f R freezes and I have to kill it with task manager when I try: dyn.load(f1.dll) Please, help. -- View this message in context: http://www.nabble.com/dll-problem-tp18448005p18448005.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org 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.
