Re: [R] readline() and Rterm in Windows
I've tried your proposal in a number of ways, and
there must be something I'm not understanding. If I
run your script (using source() in RGui, or ctrl-R
from the R Editor, I get:
conout - file('CONOUT$','w')
Error in file(CONOUT$, w) : unable to open
connection
In addition: Warning message:
cannot open file 'CONOUT$', reason 'Permission denied'
so I added the path as in:
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
print(id)
Using RGui and ctrl-R from the R Editor, I get
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
Error in file(C:\\R\\R-2.2.0\\CONIN$, r) :
unable to open connection
In addition: Warning message:
cannot open file 'C:\R\R-2.2.0\CONIN$', reason 'No
such file or directory'
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
Error in readLines(conin, 1) : object conin not
found
and with
source(foo.R)
Error in file(C:\\R\\R-2.2.0\\CONIN$, r) :
unable to open connection
In addition: Warning message:
cannot open file 'C:\R\R-2.2.0\CONIN$', reason 'No
such file or directory'
When I create a batch file with the following command
:
C:\R\R-2.2.0\bin\Rterm.exe --vanilla
C:\R\R-2.2.0\foo.R C:\R\R-2.2.0\foo.out
and double click on the batch file, the out file gives
me:
R : Copyright 2005, The R Foundation for Statistical
Computing
Version 2.2.0 (2005-10-06 r35749)
ISBN 3-900051-07-0
. . .
Type 'q()' to quit R.
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
and nothing else. In none of the situations do I get
prompted for input. What am I doing hopelessly wrong?
Mikkel
--- Duncan Murdoch [EMAIL PROTECTED] wrote:
Mikkel Grum wrote:
Duncan and Prof, thanks for your comments and
apologies for not being more specific. I'm not
getting
the same results you get from the steps you
propose.
If I write a script foo.R with two lines
id - readline(Please enter an ID: )
id
and then use source(foo.R) (either at the Rterm
prompt, or in RGui) it is true that get prompted,
but
the second line does not visibly run, i.e. I get
source(id.r)
Please enter an ID: 5
and if I then type id, I get
id
[1] id
If I cut and paste the two lines in RGui (in one
go),
I get
id - readline(Please enter an ID: )
Please enter an ID: id
What I really want is a batch file on the desktop
with
the following commands:
c:\r\R-2.2.0\bin\Rterm.exe --no-save
--no-restore
script.R script.out 21
c:\texmf\miktex\bin\latex
\nonstopmode\input{blue.tex}
and script.R reads something like:
id - readline(Please enter an ID: )
id
Sweave(blue.Rnw)
I said that script.R didn't run, which was an
incorrect description. It runs without prompting
for
the ID, and gives error messages all through
because
blue.Rnw needs the id.
This is a very simplified version of what I'm
doing,
but if I use only the first line of the batch file
and
the first two lines of the script and could get
that
to work, I could figure out the rest.
It won't work so simply. You're redirecting stdin,
so user input would
be taken from there; you're redirecting stdout and
stderr, so the prompt
won't be visible to the user.
You need to open new handles to the console. The
code below will do it
in Windows; the syntax to specify the console in
Unix-alikes will be
different (but I don't know what it is).
conout - file('CONOUT$','w')
conin - file('CONIN$', 'r')
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
print(id)
Duncan Murdoch
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] R save very huge matrices in files
I have to work with really huge matrices (about 1000*1000 or more). And I want to save those matrices in some file on my computer. I tried to do so by using the command write.tabe(SMatrix,file=C:/Programme/rw1061/SMatrix.txt,sep= ,quote=FALSE,row.names=FALSE,col.names=FALSE) SMatrix is the matrix I want as a file. Unfortunately this does not work. Error message: Error: cannot allocate vector of size 32665 Kb In addition: Warning message: Reached total allocation of 255Mb: see help(memory.size) Is there any command, which could help? What can I to in order to save SMatrix? Moritz Marienfeld __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] readline() and Rterm in Windows
Mikkel Grum wrote:
I've tried your proposal in a number of ways, and
there must be something I'm not understanding. If I
run your script (using source() in RGui, or ctrl-R
from the R Editor, I get:
It requires a command line console, i.e. it will only work in Rterm, not
Rgui. I was assuming you'd run it using the style of your batch file
down below, but without changing the paths.
Duncan Murdoch
conout - file('CONOUT$','w')
Error in file(CONOUT$, w) : unable to open
connection
In addition: Warning message:
cannot open file 'CONOUT$', reason 'Permission denied'
so I added the path as in:
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
print(id)
Using RGui and ctrl-R from the R Editor, I get
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
Error in file(C:\\R\\R-2.2.0\\CONIN$, r) :
unable to open connection
In addition: Warning message:
cannot open file 'C:\R\R-2.2.0\CONIN$', reason 'No
such file or directory'
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
Error in readLines(conin, 1) : object conin not
found
and with
source(foo.R)
Error in file(C:\\R\\R-2.2.0\\CONIN$, r) :
unable to open connection
In addition: Warning message:
cannot open file 'C:\R\R-2.2.0\CONIN$', reason 'No
such file or directory'
When I create a batch file with the following command
:
C:\R\R-2.2.0\bin\Rterm.exe --vanilla
C:\R\R-2.2.0\foo.R C:\R\R-2.2.0\foo.out
and double click on the batch file, the out file gives
me:
R : Copyright 2005, The R Foundation for Statistical
Computing
Version 2.2.0 (2005-10-06 r35749)
ISBN 3-900051-07-0
. . .
Type 'q()' to quit R.
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
and nothing else. In none of the situations do I get
prompted for input. What am I doing hopelessly wrong?
Mikkel
--- Duncan Murdoch [EMAIL PROTECTED] wrote:
Mikkel Grum wrote:
Duncan and Prof, thanks for your comments and
apologies for not being more specific. I'm not
getting
the same results you get from the steps you
propose.
If I write a script foo.R with two lines
id - readline(Please enter an ID: )
id
and then use source(foo.R) (either at the Rterm
prompt, or in RGui) it is true that get prompted,
but
the second line does not visibly run, i.e. I get
source(id.r)
Please enter an ID: 5
and if I then type id, I get
id
[1] id
If I cut and paste the two lines in RGui (in one
go),
I get
id - readline(Please enter an ID: )
Please enter an ID: id
What I really want is a batch file on the desktop
with
the following commands:
c:\r\R-2.2.0\bin\Rterm.exe --no-save
--no-restore
script.R script.out 21
c:\texmf\miktex\bin\latex
\nonstopmode\input{blue.tex}
and script.R reads something like:
id - readline(Please enter an ID: )
id
Sweave(blue.Rnw)
I said that script.R didn't run, which was an
incorrect description. It runs without prompting
for
the ID, and gives error messages all through
because
blue.Rnw needs the id.
This is a very simplified version of what I'm
doing,
but if I use only the first line of the batch file
and
the first two lines of the script and could get
that
to work, I could figure out the rest.
It won't work so simply. You're redirecting stdin,
so user input would
be taken from there; you're redirecting stdout and
stderr, so the prompt
won't be visible to the user.
You need to open new handles to the console. The
code below will do it
in Windows; the syntax to specify the console in
Unix-alikes will be
different (but I don't know what it is).
conout - file('CONOUT$','w')
conin - file('CONIN$', 'r')
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
print(id)
Duncan Murdoch
__
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com
__
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
Re: [R] RODBC and Excel: Wrong Data Type Assumed on Import
Hi As I now exclusively use copy paste method to transfer data from Excel to R I tried it and I got correctly a factor column when there were some non numeric data in Excel. Ctrl-C in Excel mydf-read.delim(clipboard) in R Are you sure that a respective column in Excel has values 275a and 275b in it? If yes I had tried to define colClasses vector for your columns. HTH Petr On 2 Nov 2005 at 12:45, Earl F. Glynn wrote: To: r-help@stat.math.ethz.ch From: Earl F. Glynn [EMAIL PROTECTED] Date sent: Wed, 2 Nov 2005 12:45:53 -0600 Subject:[R] RODBC and Excel: Wrong Data Type Assumed on Import The first column in my Excel sheet has mostly numbers but I need to treat it as character data: library(RODBC) channel - odbcConnectExcel(U:/efg/lab/R/Plasmid/construct list.xls) plasmid - sqlFetch(channel,Sheet1, as.is=TRUE) odbcClose(channel) names(plasmid) [1] Plasmid Number PlasmidConcentration Comments Lost # How is the type decided? I need a character type. class(plasmid$Plasmid Number) [1] numeric typeof(plasmid$Plasmid Number) [1] double plasmid$Plasmid Number[273:276] [1] 274 NA NA 276 The two NAs are supposed to be 275a and 275b. I tried the as.is=TRUE but that didn't help. I consulted Section 4, Relational databases, in the R Data Import/Export document (for Version 2.2.0). Section 4.2.2, Data types, was not helpful. In particular, this did not seem helpful: The more comprehensive of the R interface packages hide the type conversion issues from the user. Section 4.3.2, Package RODBC, provided a simple example of using ODBC .. with a(sic) Excel spreadsheet but is silent on how to control the data type on import. Could the documentation be expanded to address this issue? I really need to show Plasmid 275a and Plasmid 275b instead of Plasmid NA. Thanks for any help with this. efg -- Earl F. Glynn Scientific Programmer Bioinformatics Department Stowers Institute for Medical Research __ 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 Petr Pikal [EMAIL PROTECTED] __ 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
Re: [R] Bug report on get.hist.quote
get.hist.quote(instrument=INR/USD, provider=oanda, start=2005-10-20) trying URL 'http://www.oanda.com/convert/fxhistory?lang=endate1=10%2F20%2F2005date=11%2F01%2F2005date_fmt=usexch=INRexch2=expr=USDexpr2=margin_fixed=0SUBMIT=Get+Tableformat=ASCIIredirected=1' Content type 'text/html' length unknown opened URL .. ... downloaded 13Kb 2005-10-20 2005-10-21 2005-10-22 2005-10-23 2005-10-24 2005-10-25 2005-10-26 0.022200.022180.022240.022240.022240.022190.02226 2005-10-27 2005-10-28 2005-10-29 2005-10-30 2005-10-31 2005-11-01 0.022240.022250.022240.022240.022240.02221 The answer is wrong. What is shown here is USD/INR, not INR/USD. Ajay, This is not a bug! The answer is correct. *1 Indian Rupee = 0.02220 US Dollar on *2005-10-20 and this is also the standard way currencies are quoted. Pls read, e.g. http://fxtrade.oanda.com/currency_trading/fxbasics.shtml Best regards Adrian __ 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
[R] quadratic form
On page 22 of the R-introduction guide it's written:
the quadratic form x^{'} A^{-1} x which is used in
multivariate computations, should be computed by
something like x%*%solve(A,x), rather than computing
the inverse of A.
Why isn't it good to compute t(x) %*% solve(A) %*% x?
Thanks a lot for help!
__
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
[R] merging dataframes
Dear List, I often have to merge two or more data frames containing unique row names but with some columns (names) common to the two data frames and some columns not common. This toy example will explain the kind of setup I am talking about: mat1 - as.data.frame(matrix(rnorm(20), nrow = 5)) mat2 - as.data.frame(matrix(rnorm(20), nrow = 4)) rownames(mat1) - paste(site, 1:5, sep=) rownames(mat2) - paste(site, 6:9, sep=) names(mat1) - paste(species, c(1,3,5,7), sep=) names(mat2) - paste(species, c(2,3,4,7,9), sep=) mat1 mat2 So sites (rows) are unique across both data frames, but there are only 7 unique species (columns): unique(c(names(mat1), names(mat2))) merge(mat1, mat2, all = TRUE) gives almost what I want, but it drops or looses the rownames() information from the two merged data frames, and it re-orders the rows so that one simply cannot write back the correct row names. How might I go about merging two data frames as I have described, but preserve the row names and more importantly, keep the order of rows the same, so that rows from mat1 come before the rows of mat2? Many thanks, Gavin -- %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% Gavin Simpson [T] +44 (0)20 7679 5522 ENSIS Research Fellow [F] +44 (0)20 7679 7565 ENSIS Ltd. ECRC [E] gavin.simpsonATNOSPAMucl.ac.uk UCL Department of Geography [W] http://www.ucl.ac.uk/~ucfagls/cv/ 26 Bedford Way[W] http://www.ucl.ac.uk/~ucfagls/ London. WC1H 0AP. %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% __ 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
Re: [R] ML optimization question--unidimensional unfolding scalin g
Alternatively, just type debug(optim) before using it, then step through it by hitting enter repeatedly... When you're done, do undebug(optim). Andy From: Spencer Graves Have you looked at the code for optim? If you execute optim, it will list the code. You can copy that into a script file and walk through it line by line to figure out what it does. By doing this, you should be able to find a place in the iteration where you can test both branches of each bifurcation and pick one -- or keep a list of however many you want and follow them all more or less simultaneously, pruning the ones that seem too implausible. Then you can alternate between a piece of the optim code, bifurcating and pruning, adjusting each and printing intermediate progress reports to help you understand what it's doing and how you might want to modify it. With a bit more effort, you can get the official source code with comments. To do that, I think you go to www.r-project.org - CRAN - (select a local mirror) - Software: R sources. From there, just download The latest release: R-2.2.0.tar.gz. For more detailed help, I suggest you try to think of the simplest possible toy problem that still contains one of the issues you find most difficult. Then send that to this list. If readers can copy a few lines of R code from your email into R and try a couple of things in less than a minute, I think you might get more useful replies quicker. Best Wishes, Spencer Graves Peter Muhlberger wrote: Hi Spencer: Thanks for your interest! Also, the posting guide was helpful. I think my problem might be solved if I could find a way to terminate nlm or optim runs from within the user-given minimization function they call. Optimization is unconstrained. I'm essentially using normal like curves that translate observed values on a set of variables (one curve per variable) into latent unfolded values. The observed values are on the Y-axis the latent (hence parameters to be estimated) are on the X-axis. The problem is that there are two points into which an observed value can map on a curve--one on either side of the curve mean. Only one of these values actually will be optimal for all observed variables, but it's easy to show that most estimation methods will get stuck on the non-optimal value if they find that one first. Moving away from that point, the likelihood gets a whole lot worse before the routine will 'see' the optimal point on the other side of the normal curve. SANN might work, but I kind of wonder how useful it'd be in estimating hundreds of parameters--thanks to that latent scale. My (possibly harebrained) thought for how to estimate this unfolding using some gradient-based method would be to run through some iterations and then check to see whether a better solution exists on the 'other side' of the normal curves. If it does, replace those parameters with the better ones. Because this causes the likelihood to jump, I'd probably have to start the estimation process over again (maybe). But, I see no way from within the minimization function called by NLM or optim to tell NLM or optim to terminate its current run. I could make the algorithm recursive, but that eats up resources will probably have to be terminated w/ an error. Peter On 10/11/05 11:11 PM, Spencer Graves [EMAIL PROTECTED] wrote: There may be a few problems where ML (or more generally Bayes) fails to give sensible answers, but they are relatively rare. What is your likelihood? How many parameters are you trying to estimate? Are you using constrained or unconstrained optimization? If constrained, I suggest you remove the constraints by appropriate transformation. When considering alternative transformations, I consider (a) what makes physical sense, and (b) which transformation produces a log likelihood that is more close to being parabolic. Hou are you calling optim? Have you tried all SANN as well as Nelder-Mead, BFGS, and CG? If you are using constrained optimization, I suggest you move the constraints to Inf by appropriate transformation and use the other methods, as I just suggested. If you would still like more suggestions from this group, please provide more detail -- but as tersely as possible. The posting guide is, I believe, quite useful (www.R-project.org/posting-guide.html). spencer graves __ 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 -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED]
[R] How to calculate errors in histogram values
Hi there, I'm new to R but I thought this is the most likely place I could get advice or hints w.r.t the following problem: I have a series of measurements xi with associated uncertainties dxi. I would like to construct the probability density histogram of this data where each density estimate has an associated error that is derived from the dxi. In other words, for large dxi the histogram should also display large uncertainties and vice versa. I need this for a curve fitting algorithm. I have seen many crude ways of working out the error in each bin based on the bin count alone, but that's obviously independent of the dxi and thus not what I'm after. So, 1) Is there an R package that can do this (there's nothing in the refence of 2.1.1)? If so, what algorithm does it use? 2) Could anybody please point me in the right direction (papers, books, websites etc.) Thanks, -- Kilian Hagemann Climate Systems Analysis Group University of Cape Town Republic of South Africa Tel(w): ++27 21 650 2748 __ 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
Re: [R] quadratic form
Alvarez Pedro wrote:
On page 22 of the R-introduction guide it's written:
the quadratic form x^{'} A^{-1} x which is used in
multivariate computations, should be computed by
something like x%*%solve(A,x), rather than computing
the inverse of A.
Why isn't it good to compute t(x) %*% solve(A) %*% x?
Thanks a lot for help!
__
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
See
Bates, D. (2004): Least Squares Calculations in R. R News 4(1), 17-20,
http://CRAN.R-project.org/doc/Rnews/
Uwe Ligges
__
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
Re: [R] merging dataframes
On Thu, 3 Nov 2005, Gavin Simpson wrote: Dear List, I often have to merge two or more data frames containing unique row names but with some columns (names) common to the two data frames and some columns not common. This toy example will explain the kind of setup I am talking about: mat1 - as.data.frame(matrix(rnorm(20), nrow = 5)) mat2 - as.data.frame(matrix(rnorm(20), nrow = 4)) rownames(mat1) - paste(site, 1:5, sep=) rownames(mat2) - paste(site, 6:9, sep=) names(mat1) - paste(species, c(1,3,5,7), sep=) names(mat2) - paste(species, c(2,3,4,7,9), sep=) mat1 mat2 So sites (rows) are unique across both data frames, but there are only 7 unique species (columns): unique(c(names(mat1), names(mat2))) merge(mat1, mat2, all = TRUE) gives almost what I want, but it drops or looses the rownames() information from the two merged data frames, and it re-orders the rows so that one simply cannot write back the correct row names. How might I go about merging two data frames as I have described, but preserve the row names and more importantly, keep the order of rows the same, so that rows from mat1 come before the rows of mat2? merge(mat1, mat2, all = TRUE, sort=FALSE) seems to fix the second question. The first is mentioned tangentially in the help page details if you are merging on row names, which you are not - maybe prepend to both a column called sites: mat1a - data.frame(sites=row.names(mat1), mat1) mat2a - data.frame(sites=row.names(mat2), mat2) data.frame(merge(mat1a, mat2a, all = TRUE, sort=FALSE), row.names=sites) is a bit long-winded, but gets you there. Roger Many thanks, Gavin -- Roger Bivand Economic Geography Section, Department of Economics, Norwegian School of Economics and Business Administration, Helleveien 30, N-5045 Bergen, Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43 e-mail: [EMAIL PROTECTED] __ 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
Re: [R] merging dataframes
you could use something like: mat1$id1 - 1:nrow(mat1) mat2$id2 - 1:nrow(mat2) out - merge(mat1, mat2, all = TRUE) out[order(out$id1, out$id2), ] I hope it helps. Best, Dimitris Dimitris Rizopoulos Ph.D. Student Biostatistical Centre School of Public Health Catholic University of Leuven Address: Kapucijnenvoer 35, Leuven, Belgium Tel: +32/(0)16/336899 Fax: +32/(0)16/337015 Web: http://www.med.kuleuven.be/biostat/ http://www.student.kuleuven.be/~m0390867/dimitris.htm - Original Message - From: Gavin Simpson [EMAIL PROTECTED] To: R-help R-help@stat.math.ethz.ch Sent: Thursday, November 03, 2005 2:08 PM Subject: [R] merging dataframes Dear List, I often have to merge two or more data frames containing unique row names but with some columns (names) common to the two data frames and some columns not common. This toy example will explain the kind of setup I am talking about: mat1 - as.data.frame(matrix(rnorm(20), nrow = 5)) mat2 - as.data.frame(matrix(rnorm(20), nrow = 4)) rownames(mat1) - paste(site, 1:5, sep=) rownames(mat2) - paste(site, 6:9, sep=) names(mat1) - paste(species, c(1,3,5,7), sep=) names(mat2) - paste(species, c(2,3,4,7,9), sep=) mat1 mat2 So sites (rows) are unique across both data frames, but there are only 7 unique species (columns): unique(c(names(mat1), names(mat2))) merge(mat1, mat2, all = TRUE) gives almost what I want, but it drops or looses the rownames() information from the two merged data frames, and it re-orders the rows so that one simply cannot write back the correct row names. How might I go about merging two data frames as I have described, but preserve the row names and more importantly, keep the order of rows the same, so that rows from mat1 come before the rows of mat2? Many thanks, Gavin -- %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% Gavin Simpson [T] +44 (0)20 7679 5522 ENSIS Research Fellow [F] +44 (0)20 7679 7565 ENSIS Ltd. ECRC [E] gavin.simpsonATNOSPAMucl.ac.uk UCL Department of Geography [W] http://www.ucl.ac.uk/~ucfagls/cv/ 26 Bedford Way[W] http://www.ucl.ac.uk/~ucfagls/ London. WC1H 0AP. %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% __ 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 Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm __ 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
Re: [R] merging dataframes
The `Value' section of ?merge does say that `... in all cases the result has no special row names', so you're left to handle that on your own. One possibility is to use result - merge(mat1, mat2, all=TRUE, sort=FALSE) so that the sorting is not done, then you can just do rownames(result) - c(rownames(mat1), rownames(mat2)) Cheers, Andy From: Gavin Simpson Dear List, I often have to merge two or more data frames containing unique row names but with some columns (names) common to the two data frames and some columns not common. This toy example will explain the kind of setup I am talking about: mat1 - as.data.frame(matrix(rnorm(20), nrow = 5)) mat2 - as.data.frame(matrix(rnorm(20), nrow = 4)) rownames(mat1) - paste(site, 1:5, sep=) rownames(mat2) - paste(site, 6:9, sep=) names(mat1) - paste(species, c(1,3,5,7), sep=) names(mat2) - paste(species, c(2,3,4,7,9), sep=) mat1 mat2 So sites (rows) are unique across both data frames, but there are only 7 unique species (columns): unique(c(names(mat1), names(mat2))) merge(mat1, mat2, all = TRUE) gives almost what I want, but it drops or looses the rownames() information from the two merged data frames, and it re-orders the rows so that one simply cannot write back the correct row names. How might I go about merging two data frames as I have described, but preserve the row names and more importantly, keep the order of rows the same, so that rows from mat1 come before the rows of mat2? Many thanks, Gavin -- %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~ %~%~%~%~% Gavin Simpson [T] +44 (0)20 7679 5522 ENSIS Research Fellow [F] +44 (0)20 7679 7565 ENSIS Ltd. ECRC [E] gavin.simpsonATNOSPAMucl.ac.uk UCL Department of Geography [W] http://www.ucl.ac.uk/~ucfagls/cv/ 26 Bedford Way [W] http://www.ucl.ac.uk/~ucfagls/ London. WC1H 0AP. %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~ %~%~%~%~% __ 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 __ 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
Re: [R] quadratic form
On Thu, 3 Nov 2005, Alvarez Pedro wrote:
On page 22 of the R-introduction guide it's written:
the quadratic form x^{'} A^{-1} x which is used in
multivariate computations, should be computed by
something like x%*%solve(A,x), rather than computing
the inverse of A.
Why isn't it good to compute t(x) %*% solve(A) %*% x?
The answer is only two lines above;
Numerically, it is both inefficient and potentially unstable to compute
@code{x - solve(A) %*% b} instead of @code{solve(A,b)}.
See also the footnote.
--
Brian D. Ripley, [EMAIL PROTECTED]
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@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
Re: [R] readline() and Rterm in Windows
And that was the only combination I didn't try, duhh.
As you say, it works. Excellent! TX! They'll be a
number of pleased usres too.
--- Duncan Murdoch [EMAIL PROTECTED] wrote:
Mikkel Grum wrote:
I've tried your proposal in a number of ways, and
there must be something I'm not understanding. If
I
run your script (using source() in RGui, or ctrl-R
from the R Editor, I get:
It requires a command line console, i.e. it will
only work in Rterm, not
Rgui. I was assuming you'd run it using the style
of your batch file
down below, but without changing the paths.
Duncan Murdoch
conout - file('CONOUT$','w')
Error in file(CONOUT$, w) : unable to open
connection
In addition: Warning message:
cannot open file 'CONOUT$', reason 'Permission
denied'
so I added the path as in:
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
print(id)
Using RGui and ctrl-R from the R Editor, I get
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
Error in file(C:\\R\\R-2.2.0\\CONIN$, r) :
unable to open connection
In addition: Warning message:
cannot open file 'C:\R\R-2.2.0\CONIN$', reason 'No
such file or directory'
cat('Please enter an ID:', file=conout)
flush(conout)
id - readLines(conin, 1)
Error in readLines(conin, 1) : object conin not
found
and with
source(foo.R)
Error in file(C:\\R\\R-2.2.0\\CONIN$, r) :
unable to open connection
In addition: Warning message:
cannot open file 'C:\R\R-2.2.0\CONIN$', reason 'No
such file or directory'
When I create a batch file with the following
command
:
C:\R\R-2.2.0\bin\Rterm.exe --vanilla
C:\R\R-2.2.0\foo.R C:\R\R-2.2.0\foo.out
and double click on the batch file, the out file
gives
me:
R : Copyright 2005, The R Foundation for
Statistical
Computing
Version 2.2.0 (2005-10-06 r35749)
ISBN 3-900051-07-0
. . .
Type 'q()' to quit R.
conout - file('C:\\R\\R-2.2.0\\CONOUT$','w')
conin - file('C:\\R\\R-2.2.0\\CONIN$', 'r')
and nothing else. In none of the situations do I
get
prompted for input. What am I doing hopelessly
wrong?
Mikkel
--- Duncan Murdoch [EMAIL PROTECTED] wrote:
Mikkel Grum wrote:
Duncan and Prof, thanks for your comments and
apologies for not being more specific. I'm not
getting
the same results you get from the steps you
propose.
If I write a script foo.R with two lines
id - readline(Please enter an ID: )
id
and then use source(foo.R) (either at the Rterm
prompt, or in RGui) it is true that get prompted,
but
the second line does not visibly run, i.e. I get
source(id.r)
Please enter an ID: 5
and if I then type id, I get
id
[1] id
If I cut and paste the two lines in RGui (in one
go),
I get
id - readline(Please enter an ID: )
Please enter an ID: id
What I really want is a batch file on the desktop
with
the following commands:
c:\r\R-2.2.0\bin\Rterm.exe --no-save
--no-restore
script.R script.out 21
c:\texmf\miktex\bin\latex
\nonstopmode\input{blue.tex}
and script.R reads something like:
id - readline(Please enter an ID: )
id
Sweave(blue.Rnw)
I said that script.R didn't run, which was an
incorrect description. It runs without prompting
for
the ID, and gives error messages all through
because
blue.Rnw needs the id.
This is a very simplified version of what I'm
doing,
but if I use only the first line of the batch
file
and
the first two lines of the script and could get
that
to work, I could figure out the rest.
It won't work so simply. You're redirecting
stdin,
so user input would
be taken from there; you're redirecting stdout and
stderr, so the prompt
won't be visible to the user.
=== message truncated ===
__
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
[R] merging dataframes
Hi, what about padding both datasets with dummy missing records ... and then play with cbind and rbind ... like e.g. species5-c(NA,NA,NA,NA) modmat2-cbind(mat2,species1,species5) and then similarly with mat1 ... e.g. species2-c(NA,NA,NA,NA,NA) modmad1-cbind(mat1,species2,species4,species9) rbind(modmad1,modmat2) species1species3 species5 species7 species2 species4 species9 site1 -0.7190044 -0.52482580 -1.1813567 -1.5584831 NA NA NA site2 -1.1782180 1.72337964 0.1652343 -0.9026087 NA NA NA site3 0.3823015 -0.07226644 -1.2907470 -0.3692091 NA NA NA site4 -1.3051131 -0.61107947 0.6264416 1.5259373 NA NA NA site5 0.2028565 -1.28374638 1.6284780 -1.2975163 NA NA NA site6 NA 1.19088414 NA 0.3159949 -0.1624538 0.5987733 0.2205512 site7 NA 0.75292176 NA 1.7524988 0.8335334 -0.7998774 -0.9788762 site8 NA -0.47803396 NA -1.3041628 1.7925165 -0.4153879 -0.4708165 site9 NA -0.20063523 NA 1.8119115 1.5351801 -1.3334419 0.5812675 it will need modifications of course if you are working with several datasets Saludos, Manuel __ 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
Re: [R] merging dataframes
On Thu, 2005-11-03 at 08:33 -0500, Liaw, Andy wrote:
The `Value' section of ?merge does say that `... in all cases the result has
no special row names', so you're left to handle that on your own. One
possibility is to use
result - merge(mat1, mat2, all=TRUE, sort=FALSE)
so that the sorting is not done, then you can just do
rownames(result) - c(rownames(mat1), rownames(mat2))
Cheers,
Andy
Thanks Roger, Andy and Dimitris for your solutions.
I'd missed the default for sort being set to TRUE - must pay more
attention in class. Roger's and Andy's approaches seem like the most
fool-proof way of canning this into a function that does all the I asked
and a bit of tidying up of NA's etc.
Cheers,
G
From: Gavin Simpson
Dear List,
I often have to merge two or more data frames containing unique row
names but with some columns (names) common to the two data frames and
some columns not common. This toy example will explain the
kind of setup
I am talking about:
mat1 - as.data.frame(matrix(rnorm(20), nrow = 5))
mat2 - as.data.frame(matrix(rnorm(20), nrow = 4))
rownames(mat1) - paste(site, 1:5, sep=)
rownames(mat2) - paste(site, 6:9, sep=)
names(mat1) - paste(species, c(1,3,5,7), sep=)
names(mat2) - paste(species, c(2,3,4,7,9), sep=)
mat1
mat2
So sites (rows) are unique across both data frames, but there
are only 7
unique species (columns):
unique(c(names(mat1), names(mat2)))
merge(mat1, mat2, all = TRUE)
gives almost what I want, but it drops or looses the rownames()
information from the two merged data frames, and it re-orders the rows
so that one simply cannot write back the correct row names.
How might I go about merging two data frames as I have described, but
preserve the row names and more importantly, keep the order
of rows the
same, so that rows from mat1 come before the rows of mat2?
Many thanks,
Gavin
--
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~
%~%~%~%~%
Gavin Simpson [T] +44 (0)20 7679 5522
ENSIS Research Fellow [F] +44 (0)20 7679 7565
ENSIS Ltd. ECRC [E] gavin.simpsonATNOSPAMucl.ac.uk
UCL Department of Geography [W]
http://www.ucl.ac.uk/~ucfagls/cv/
26 Bedford Way
[W] http://www.ucl.ac.uk/~ucfagls/
London. WC1H 0AP.
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~
%~%~%~%~%
__
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
--
Notice: This e-mail message, together with any attachment...{{dropped}}
__
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
[R] Potential for R to conflict with other softwares
Hi. After some time, my collegues at the Food and Drug Adminstration have finally acknowledged R as a powerful statistical computing environment. However, in order to comply with the Office of Information and Technology standards there are a couple of questions about whether R could interfere with other software. As I'm more of a driver of the R software and not a mechanic, I was hoping for the insight of the many great useRs. Below is a list of 5 proposed questions to which I value any comment. Thank you for your time, -Mat 1. Does R have high resolution graphics? 2. Does R have .dll files, or other executables which are not located in the R software directory tree? 3. Does R modify the Windows registry in a non-obvious way, i.e. other than defining itself and what extensions to associate with R, and what are those extensions? 4. Does R add macros to any part of MS Office? 5. Can you anticipate any other way in which installing and using R could disrupt the operation of another software? *** Mat Soukup, Ph.D. Food and Drug Administration 10903 New Hampshire Ave. BLDG 22 RM 5329 Silver Spring, MD 20993-0002 Phone: 301.796.1005 *** [[alternative HTML version deleted]] __ 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
[R] problems with pan(): Indizierung ausserhalb der Grenzen = subscript out of bounds
Dear alltogether, I tried pan() to impute NAs for longitudinal data. The terminology in the following output follows the pan manpage. No data are attached to this script as this may be too huge. y = 15 responses pred = at first just intercept was tried (later on covariates should follow) subj = 168 different subjects with 4 to 6 observations for each subject at time points t1, t2, ..., t6 # extract of y y[1:4,] anpr impr kepr lernpr lstofpr nachwela nachfaknachw sbpr widapr zdompr zerfpr zgleipr zstimpr zugrupr 2 3.50 2.75 3.4 2.22 2.67 3.33 15.00 5.909091 2.33 21.5 3.67 4 3.00 3.56 202 2.25 2.50 3.6 2.22 3.67 12.00 13.75 7.78 1.67 22.0 3.33 4 3.33 3.56 402 NA NA NA NA NANANA NA NA NA NA NA NA NA NA 602 1.75 2.75 3.4 1.56 3.33 2.67 6.67 5.00 2.00 21.0 3.33 4 2.67 3.33 dim(y) [1] 940 15 # matrix y with 15 responses and 940 obs # y is ordered according to subj length(subj) [1] 940 #940 obs of 168 different subjects (persons) # extract of subj subj[1:30] [1] 2 2 2 2 2 2 3 3 3 3 3 4 4 4 4 4 5 5 5 5 5 5 6 6 6 6 6 6 7 7 # how many observations for each subject table(subj) subj 2 3 4 5 6 7 8 9 12 13 14 15 16 17 18 19 20 21 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 41 42 43 44 46 47 48 49 6 5 5 6 6 6 5 6 6 6 6 6 5 6 6 5 6 6 6 6 5 6 6 5 6 6 4 6 5 5 6 6 6 5 6 6 4 6 5 6 5 6 6 50 52 53 55 57 58 59 60 61 62 64 66 67 68 69 70 72 73 75 76 77 78 79 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 97 98 99 100 101 6 5 5 5 6 6 5 5 6 6 6 6 4 6 6 5 5 6 6 5 6 4 6 5 5 4 6 5 6 6 4 5 5 6 6 6 6 6 6 6 6 6 6 103 104 105 106 107 108 109 110 112 113 114 115 116 117 118 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 137 138 139 140 141 142 144 146 147 148 149 150 151 5 6 5 6 6 6 5 5 5 6 6 5 5 6 5 6 6 6 6 6 5 4 6 6 6 6 6 4 4 6 5 4 6 5 4 6 6 6 6 6 6 5 5 152 153 154 155 156 157 158 159 160 162 164 165 166 167 170 171 173 174 175 176 178 179 181 182 185 186 187 188 189 190 191 192 193 195 196 197 198 199 200 5 6 6 6 5 6 6 6 6 6 6 6 6 4 5 6 6 6 6 5 6 6 6 6 6 6 6 5 6 6 6 5 6 6 6 6 6 6 6 pred - cbind(interc=rep(1,dim(y)[1])) # just intercept (at first) dim(pred) [1] 940 1 xcol - 1:dim(pred)[2] xcol [1] 1 #xcol = 1 , using all number of cols of pred[] zcol - c(1)# = 1 , number of cols to use y.ncol - dim(y)[2] n.zcol - length(zcol) prior - list(a=y.ncol, + Binv=diag(y.ncol), + c=n.zcol, + Dinv=diag(n.zcol)) prior $a [1] 15 $Binv [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] [,14] [,15] [1,]100000000 0 0 0 0 0 0 [2,]010000000 0 0 0 0 0 0 [3,]001000000 0 0 0 0 0 0 [4,]000100000 0 0 0 0 0 0 [5,]000010000 0 0 0 0 0 0 [6,]000001000 0 0 0 0 0 0 [7,]000000100 0 0 0 0 0 0 [8,]000000010 0 0 0 0 0 0 [9,]000000001 0 0 0 0 0 0 [10,]000000000 1 0 0 0 0 0 [11,]000000000 0 1 0 0 0 0 [12,]000000000 0 0 1 0 0 0 [13,]000000000 0 0 0 1 0 0 [14,]000000000 0 0 0 0 1 0 [15,]000000000 0 0 0 0 0 1 $c [1] 1 $Dinv [,1] [1,]1 #prior a = number of cols in y # Binv = identity matrix (ncols = nrows = y) # c = length of zcol[] # Dinv = identity matrix (ncols = length(nzcol)) Now the error message: pan(y, subj, pred, xcol, zcol, prior, seed=1234,iter=1000) Fehler: Indizierung außerhalb der Grenzen # error massage = subscript out of bounds I do not understand that. Thank you very
Re: [R] quadratic form
Alvarez Pedro [EMAIL PROTECTED] writes:
On page 22 of the R-introduction guide it's written:
the quadratic form x^{'} A^{-1} x which is used in
multivariate computations, should be computed by
something like x%*%solve(A,x), rather than computing
the inverse of A.
Why isn't it good to compute t(x) %*% solve(A) %*% x?
It's just a waste of CPU time. For k x k matrices, solution of Ax=b is
of computational complexity O(k^2) whereas inversion of A is O(k^3).
This is obviously more important for k=1000 than for k=5.
--
O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B
c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K
(*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918
~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907
__
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
Re: [R] Potential for R to conflict with other softwares
On 11/3/2005 9:11 AM, Soukup, Mat wrote: Hi. After some time, my collegues at the Food and Drug Adminstration have finally acknowledged R as a powerful statistical computing environment. However, in order to comply with the Office of Information and Technology standards there are a couple of questions about whether R could interfere with other software. As I'm more of a driver of the R software and not a mechanic, I was hoping for the insight of the many great useRs. Below is a list of 5 proposed questions to which I value any comment. Thank you for your time, -Mat These answers are about the Windows version only, but from the questions, I think that's what you were looking for. They apply to all versions since 1.6.x at least (though the earlier ones would have put fewer entries into the registry, they put them in the same places). 1. Does R have high resolution graphics? Yes, but I don't think I get the point of this question. How would that interfere with other software? 2. Does R have .dll files, or other executables which are not located in the R software directory tree? No, it installs everything below R_HOME. 3. Does R modify the Windows registry in a non-obvious way, i.e. other than defining itself and what extensions to associate with R, and what are those extensions? I think all of its modifications would count as obvious. They are mainly below HKLM/Software/R-core or HKCU/Software/R-core (where the file locations are recorded); additionally file associations are set up for .Rdata files (which are called RWorkspace files there), and an uninstall entry is made. 4. Does R add macros to any part of MS Office? No. 5. Can you anticipate any other way in which installing and using R could disrupt the operation of another software? No, not really. Maybe users will become addicted to it? ;-) Duncan Murdoch *** Mat Soukup, Ph.D. Food and Drug Administration 10903 New Hampshire Ave. BLDG 22 RM 5329 Silver Spring, MD 20993-0002 Phone: 301.796.1005 *** [[alternative HTML version deleted]] __ 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 __ 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
[R] Problems with abline adding regression line to a graph
Hello all,
R2.1.1, W2k
I try to make a plot of a simple regression model in this way:
with(njfA_bcd, {
+ plot(TC_OS.G31,Prot,cex = 2, col = red, xlab= TC/OS at GS32,
+ ylab=Grain crude protein (CP))
+ })
This part works well and produces the datapoints as red circles.
When I try to add a line, using a fitted linear model in a way
that works perfect with other variables in the same dateset the
following happens:
with(njfA_bcd, {
+abline(lm(predict(m1tc) ~ TC_OS.G31), lty = 1, col = red)
+ })
Error in model.frame(formula, rownames, variables, varnames, extras,
extranames, :
variable lengths differ
And this means?
There exists missing values for TC_OS.G31 in the dataset. From the
beginning
m1tc was a lm() object, which gave the same Error message. To try to fix the
problem I changed to lme() and used na.action=na.omit explicitely, but this
didn´t help.
Here is the summary of m1tc:
summary(m1tc)
Linear mixed-effects model fit by REML
Data: njfA_bcd
AIC BIClogLik
209.4914 219.0692 -100.7457
Random effects:
Formula: ~1 | Trial
(Intercept) Residual
StdDev:1.242184 0.6520464
Fixed effects: Prot ~ TC_OS.G31
Value Std.Error DF t-value p-value
(Intercept) 14.86209 0.957630 68 15.519662 0
TC_OS.G31 -24.22286 4.792801 68 -5.054008 0
Correlation:
(Intr)
TC_OS.G31 -0.935
Standardized Within-Group Residuals:
Min Q1 Med Q3 Max
-1.68329774 -0.73751040 -0.05600477 0.68301243 2.21693174
Number of Observations: 83
Number of Groups: 14
What is happening and what shall I do about it?
Cheers
/CG
--
CG Pettersson, MSci, PhD Stud.
Swedish University of Agricultural Sciences (SLU)
Dept. of Crop Production Ecology. Box 7043.
SE-750 07 UPPSALA, Sweden.
+46 18 671428, +46 70 3306685
[EMAIL PROTECTED]
__
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
[R] Rserve/Python
Has anyone done anything on a Python client for Rserve? Simon Urbanek (Rserve dev) tells me he heard of some people working on it a couple of years ago but nothing came of it. If anyone has done anything, or might find it interesting, please get in touch with me. I know there's also the RSPython package but I'm tied to a particular python version and I can't be sure I'll get RSPython working (in Windows as well as Linux). At least if I write a Python-Rserve client its my job to make it work! Thanks, Barry Rowlingson Maths and Stats Lancaster University __ 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
Re: [R] locfit: simultaneous confidence band
Apologies for coming to this so late... Variance is rarely known in real life data. You should really consult the book `Local Regression and Likelihood' by Prof. Loader for the details on simultaneous confidence bands. `Locfit' is the support software for that book. Andy From: Michael Gälger I'm using the package 'locfit' for nonparametric regression. This package contains the function 'scb' to compute simultaneous confidence bands. The variance of the data is unknown. Up to now I compute a fit with 'locfit'. Afterwards an estimate of the residual variance is computed by the function 'rv'. The weights in the 'scb'-function are set 1/sigma^2 to compute the confidence band. Is this procedure correct or is there any other way to compute confidence bands with unknown variance? Thanks very much for any help you can offer. Michael Gälger __ 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 __ 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
[R] Search within a file
Hi, I am looking for a way to search a file for position of some expression, from within R. My current code: sha1Pos = gregexpr(sha1, readChar(filename, file.info(filename)$size))[[1]] Works fine for small files, but text files I will be working with might get up to Gb range, so I was trying to accomplish the same without loading the whole file into R. I realize this is not what R is designed to do, but maybe there is some way I am missing. Jarek Tuszynski [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Search within a file
Would this be ok (on Windows, use grep on UNIX): # line numbers of all lines conitaining R in the R README file setwd(R.home()) as.numeric(sub(:.*, , system(findstr /n R README, intern = TRUE))) On 11/3/05, Tuszynski, Jaroslaw W. [EMAIL PROTECTED] wrote: Hi, I am looking for a way to search a file for position of some expression, from within R. My current code: sha1Pos = gregexpr(sha1, readChar(filename, file.info(filename)$size))[[1]] Works fine for small files, but text files I will be working with might get up to Gb range, so I was trying to accomplish the same without loading the whole file into R. I realize this is not what R is designed to do, but maybe there is some way I am missing. Jarek Tuszynski [[alternative HTML version deleted]] __ 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 __ 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
Re: [R] npmc package
I just downloaded the file from that location, and took a quick look. At
least the only thing that R CMD check under 2.2.0 only complained about the
file report.Rd. The problem seems to be
\keyword { print }
instead of
\keyword{print}
If no one else is aware of any other problems with the package (e.g., ones
that R CMD check can not test for), please let me know. Otherwise I'll take
over the maintanence and submit it to CRAN.
Andy
From: Prof Brian Ripley
On Fri, 21 Oct 2005, Kjetil Holuerson wrote:
Martin Maechler wrote:
Carlos == Carlos Mauricio Cardeal Mendes [EMAIL PROTECTED]
on Wed, 19 Oct 2005 15:11:32 -0300 writes:
Carlos So, is there another package to substitute those
Carlos functions described on ORPHANED npmc package ?
May be not.
But nobody stops you from becoming the new maintainer of the
Just checked. This package is not now in the ORPHANES
subdirectory, neither in the main CRAN listing.
See http://cran.r-project.org/src/contrib/Archive/N/
The Orphaned (sic) subdirectory applies to packages dropped
by the former
maintainer which no longer pass R CMD check, and also to
those where the
CRAN maintainers are able to deduce it has been dropped.
Others which
just fail without a positive indication of no active
maintainer may end in
the archive.
--
Brian D. Ripley, [EMAIL PROTECTED]
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@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
__
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
Re: [R] quadratic form
Hi Alvarez
If you define
quad.form.inv - function (M, x)
{
drop(crossprod(x, solve(M, x)))
}
then you will avoid an expensive call to %*% as well.
HTH
Robin
On 3 Nov 2005, at 13:01, Alvarez Pedro wrote:
On page 22 of the R-introduction guide it's written:
the quadratic form x^{'} A^{-1} x which is used in
multivariate computations, should be computed by
something like x%*%solve(A,x), rather than computing
the inverse of A.
Why isn't it good to compute t(x) %*% solve(A) %*% x?
Thanks a lot for help!
__
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
--
Robin Hankin
Uncertainty Analyst
National Oceanography Centre, Southampton
European Way, Southampton SO14 3ZH, UK
tel 023-8059-7743
__
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
Re: [R] RODBC and Excel: Wrong Data Type Assumed on Import
From my experience (somewhat of a guess): 1. Excel uses the first 16 rows of data to determine if a column is numeric or character. The data type which is most common in the first 16 rows will then be used for the whole column. If you sort the data so that at least the first 9 rows have character data, you may find this allows the data to be interpreted as character. There is supposedly a registy setting that can control how many lines to use (instead of 16), but I have not had success with the setting. I suspect that ODBC uses JET4, which may be the real source of the problem. See more here: http://www.dicks-blog.com/archives/2004/06/03/external-data-mixed-data-types/ 2. The gregmisc bundle has a different read.xls function that uses a Perl script (xls2csv) and seems to be safer with mixed-type columns. Requires a working version of Perl. Best, Kevin Wright The first column in my Excel sheet has mostly numbers but I need to treat it as character data: library(RODBC) http://tolstoy.newcastle.edu.au/R/help/05/09/11324.html#14938qlink1 * channel - odbcConnectExcel(U:/efg/lab/R/Plasmid/construct list.xls) * * plasmid - sqlFetch(channel,Sheet1, as.is=TRUE) * * odbcClose(channel) * names(plasmid) [1] Plasmid Number Plasmid Concentration Comments Lost # How is the type decided? I need a character type. class(plasmid$Plasmid Number) [1] numeric typeof(plasmid$Plasmid Number) [1] double plasmid$Plasmid Number[273:276] [1] 274 NA NA 276 The two NAs are supposed to be 275a and 275b. I tried the as.is=TRUE but that didn't help. I consulted Section 4, Relational databases, in the R Data Import/Export document (for Version 2.2.0). Section 4.2.2, Data types, was not helpful. In particular, this did not seem helpful: The more comprehensive of the R interface packages hide the type conversion issues from the user. Section 4.3.2, Package RODBC, provided a simple example of using ODBC .. with a(sic) Excel spreadsheet but is silent on how to control the data type on import. Could the documentation be expanded to address this issue? I really need to show Plasmid 275a and Plasmid 275b instead of Plasmid NA. Thanks for any help with this. efg -- Earl F. Glynn Scientific Programmer Bioinformatics Department [[alternative HTML version deleted]] __ 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
Re: [R] How to calculate errors in histogram values
On 03-Nov-05 Kilian Hagemann wrote: Hi there, I'm new to R but I thought this is the most likely place I could get advice or hints w.r.t the following problem: I have a series of measurements xi with associated uncertainties dxi. I would like to construct the probability density histogram of this data where each density estimate has an associated error that is derived from the dxi. In other words, for large dxi the histogram should also display large uncertainties and vice versa. I need this for a curve fitting algorithm. I have seen many crude ways of working out the error in each bin based on the bin count alone, but that's obviously independent of the dxi and thus not what I'm after. So, 1) Is there an R package that can do this (there's nothing in the refence of 2.1.1)? If so, what algorithm does it use? 2) Could anybody please point me in the right direction (papers, books, websites etc.) Thanks, -- Kilian Hagemann I don't know about an R package that would deal with this directly, but I can think of an aproach, not difficult to implement in R, which may be helpful. I'm going to assume (at any rate for the time being), that you are interested in a per-bin uncertainty, i.e. that you want to be able to to answer For each bin, what is the uncertainty in the count for this bin, regardless of any other bins? I.e. you are ignoring the fact that there is correlation between bins (what goes into one bin can not go into another). 1. Say you have N observations (i = 1:N). Draw a preliminary histogram, and from this decide on a good set of fixed breaks. 2. Extend this list by a few bins on either side (you may have to return to this point, depending on the outcome of later stages). Say this gives you K bins (j = 1:K). 3. For each data-point xi, with associated dxi, and for each bin j, use this to compute the probability pij that a point with mean xi and error dxi should fall into bin j. This might be based on something as naive as integrating a Normal distribution with mean xi and SD dxi over the range of the bin j. 4. You then have an array P = p[i,j], say, where in row i you have the computed probabilities for bins 1:K 5. Now: for bin j, you have an N-column of pij values. The expected number of the N which might really be in bin j is then Ej = sum(P[,j]) and its variance (assuming that the errors in the xi are independent of each other) is Vj = sum(P[,j]*(1 - P[,j])) 6. So now you have the bin that might have been, with expected value Ej and standard deviation Sj = sqrt(Vj). Now draw a histogram (you can use 'lines()' for this in R) with bin heights Ej, and errors bars +/- Sj. It is at this stage that you may have to go back to stage 2. In order to be sure that you will not overlook xi values that spill outside the range of the bins you chose in stage 2, you need to verify that the bins you are using extend beyond the range of the original data, and that the two end-bins have negligible E1 and EK. 7. NOTE that the Ej will in general be different from the counts in bin j from the original data. This is due to overspill: if you have an original bin with a small count, which has next to it a bin with a large count, the uncertainty about whether some of the latter should really be in the former will contribute positively to the bin with the small count, by a larger amount than the bin with the small count will contribute to the bin with the large count. As you can see, this will have an effect of somewhat flattening the histogram, and of smoothing irregular variation from bin to bin. This is just an outline of a possible approach, which you may be able to develop to better suit your purposes if they are different from what I've been assuming. Best wishes, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 03-Nov-05 Time: 15:45:35 -- XFMail -- __ 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
Re: [R] Potential for R to conflict with other softwares
Duncan Murdoch [EMAIL PROTECTED] writes: On 11/3/2005 9:11 AM, Soukup, Mat wrote: Hi. After some time, my collegues at the Food and Drug Adminstration have finally acknowledged R as a powerful statistical computing environment. However, in order to comply with the Office of Information and Technology standards there are a couple of questions about whether R could interfere with other software. As I'm more of a driver of the R software and not a mechanic, I was hoping for the insight of the many great useRs. Below is a list of 5 proposed questions to which I value any comment. Thank you for your time, -Mat These answers are about the Windows version only, but from the questions, I think that's what you were looking for. They apply to all versions since 1.6.x at least (though the earlier ones would have put fewer entries into the registry, they put them in the same places). 1. Does R have high resolution graphics? Yes, but I don't think I get the point of this question. How would that interfere with other software? Device drivers! We have seen cases where the FPU control word had to be reset (DM will know this more precisely than me, I think). We do have code in place to catch that particular issue though. If you want to be really sure, I believe that batch runs using the postscript() or pdf() drivers would be immune to such issues (right?). (BTW, does the FDA trust SAS for Windows? I've seen a weird thing or two happening to the display there...) 2. Does R have .dll files, or other executables which are not located in the R software directory tree? No, it installs everything below R_HOME. We do rely on MCVCRT.DLL and some other system DLLs though, and I think this can get modified by other software. 3. Does R modify the Windows registry in a non-obvious way, i.e. other than defining itself and what extensions to associate with R, and what are those extensions? I think all of its modifications would count as obvious. They are mainly below HKLM/Software/R-core or HKCU/Software/R-core (where the file locations are recorded); additionally file associations are set up for .Rdata files (which are called RWorkspace files there), and an uninstall entry is made. 4. Does R add macros to any part of MS Office? No. 5. Can you anticipate any other way in which installing and using R could disrupt the operation of another software? No, not really. Maybe users will become addicted to it? ;-) Duncan Murdoch *** Mat Soukup, Ph.D. Food and Drug Administration 10903 New Hampshire Ave. BLDG 22 RM 5329 Silver Spring, MD 20993-0002 Phone: 301.796.1005 *** ( I think Mat is owed a big Thank you for his efforts). -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ 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
[R] MDS: Sample in one group appears twice. Why?
Hi All, I am trying to apply MDS for 4 groups in my data. The groups are: groups = list( Day1C=c(9), Day1T=c(7,8,10), Day2C=c(1,2,3,6,11,13,14,15), Day2T=c(4,5,12,16,17,18) ) When I do the MDS plot the group1 member appears twice instead of one time in the plot. I don't know why this is happening. I would greatly appreciate your help in fixing this problem. Thanks in Advance. Ezhil __ 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
[R] Help on model selection using AICc
Hi, I'm fitting poisson regression models to counts of birds in 1x1 km squares using several environmental variables as predictors. I do this in a stepwise way, using the stepAIC function. However the resulting models appear to be overparametrized, since too much variables were included. I would like to know if there is the possibility of fitting models by steps but using the AICc instead of AIC. Or at least I wonder if it would be possible to save the AIC value and number of parameters of models fitted in each step and to calculate AICc afterward. Help on this will be very much appreciated German Lopez __ 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
Re: [R] ML optimization question--unidimensional unfolding scalin g
Hi, Andy and Peter: That's interesting. I still like the idea of making my own local copy, because I can more easily add comments and test ideas while working through the code. I haven't used debug, but I think I should try it, because some things occur when running a function that don't occur when I walk through it line by line, e.g., parsing the call and ... arguments. Two more comments on the original question: 1. What is the structure of your data? Have you considered techniques for Multidimensional Scaling (MDS)? It seems that your problem is just a univariate analogue of the MDS problem. For metric MDS from a complete distance matrix, the solution is relatively straightforward computation of eigenvalues and vectors from a matrix computed from the distance matrix, and there is software widely available for the nonmetric MDS problem. For a terse introduction to that literature, see Venables and Ripley (2002) Modern Applied Statistics with S, 4th ed. (Springer, distance methods in sec. 11.1, pp. 306-308). 2. If you don't have a complete distance matrix, might it be feasible to approach the problem starting small and building larger, i.e., start with 3 nodes, then add a fourth, etc.? spencer graves Liaw, Andy wrote: Alternatively, just type debug(optim) before using it, then step through it by hitting enter repeatedly... When you're done, do undebug(optim). Andy From: Spencer Graves Have you looked at the code for optim? If you execute optim, it will list the code. You can copy that into a script file and walk through it line by line to figure out what it does. By doing this, you should be able to find a place in the iteration where you can test both branches of each bifurcation and pick one -- or keep a list of however many you want and follow them all more or less simultaneously, pruning the ones that seem too implausible. Then you can alternate between a piece of the optim code, bifurcating and pruning, adjusting each and printing intermediate progress reports to help you understand what it's doing and how you might want to modify it. With a bit more effort, you can get the official source code with comments. To do that, I think you go to www.r-project.org - CRAN - (select a local mirror) - Software: R sources. From there, just download The latest release: R-2.2.0.tar.gz. For more detailed help, I suggest you try to think of the simplest possible toy problem that still contains one of the issues you find most difficult. Then send that to this list. If readers can copy a few lines of R code from your email into R and try a couple of things in less than a minute, I think you might get more useful replies quicker. Best Wishes, Spencer Graves Peter Muhlberger wrote: Hi Spencer: Thanks for your interest! Also, the posting guide was helpful. I think my problem might be solved if I could find a way to terminate nlm or optim runs from within the user-given minimization function they call. Optimization is unconstrained. I'm essentially using normal like curves that translate observed values on a set of variables (one curve per variable) into latent unfolded values. The observed values are on the Y-axis the latent (hence parameters to be estimated) are on the X-axis. The problem is that there are two points into which an observed value can map on a curve--one on either side of the curve mean. Only one of these values actually will be optimal for all observed variables, but it's easy to show that most estimation methods will get stuck on the non-optimal value if they find that one first. Moving away from that point, the likelihood gets a whole lot worse before the routine will 'see' the optimal point on the other side of the normal curve. SANN might work, but I kind of wonder how useful it'd be in estimating hundreds of parameters--thanks to that latent scale. My (possibly harebrained) thought for how to estimate this unfolding using some gradient-based method would be to run through some iterations and then check to see whether a better solution exists on the 'other side' of the normal curves. If it does, replace those parameters with the better ones. Because this causes the likelihood to jump, I'd probably have to start the estimation process over again (maybe). But, I see no way from within the minimization function called by NLM or optim to tell NLM or optim to terminate its current run. I could make the algorithm recursive, but that eats up resources will probably have to be terminated w/ an error. Peter On 10/11/05 11:11 PM, Spencer Graves [EMAIL PROTECTED] wrote: There may be a few problems where ML (or more generally Bayes) fails to give sensible answers, but they are relatively rare. What is your likelihood? How many parameters are you trying
[R] fatal error unused tempdir
Hello, I am running R on XP Windows machine, and frequently I have enable to start R. I have this message: Fatal Error: cannot find unused tempdir name. I don't know why. Thanks, in advance, for your help Sandrine Coelho __ 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
[R] PL2006 Program Ballot
Anyone interested in voting for R to be included in the annual Pricelessware list (Windows freeware list voted on by the readers of the alt.comp.freeware newsgroup, see www.pricelesswarehome.org) can reply to this message: http://groups.google.com/group/alt.comp.freeware/msg/d9e52e406d9e2ceb simply deleting all programs you don't want to vote for and leaving R (and any other program) you wish to vote for. Deadline for voting is November 7. __ 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
Re: [R] Potential for R to conflict with other softwares
On 11/3/2005 10:46 AM, Peter Dalgaard wrote: Duncan Murdoch [EMAIL PROTECTED] writes: On 11/3/2005 9:11 AM, Soukup, Mat wrote: Hi. After some time, my collegues at the Food and Drug Adminstration have finally acknowledged R as a powerful statistical computing environment. However, in order to comply with the Office of Information and Technology standards there are a couple of questions about whether R could interfere with other software. As I'm more of a driver of the R software and not a mechanic, I was hoping for the insight of the many great useRs. Below is a list of 5 proposed questions to which I value any comment. Thank you for your time, -Mat These answers are about the Windows version only, but from the questions, I think that's what you were looking for. They apply to all versions since 1.6.x at least (though the earlier ones would have put fewer entries into the registry, they put them in the same places). 1. Does R have high resolution graphics? Yes, but I don't think I get the point of this question. How would that interfere with other software? Device drivers! We have seen cases where the FPU control word had to be reset (DM will know this more precisely than me, I think). We do have code in place to catch that particular issue though. If you want to be really sure, I believe that batch runs using the postscript() or pdf() drivers would be immune to such issues (right?). (BTW, does the FDA trust SAS for Windows? I've seen a weird thing or two happening to the display there...) 2. Does R have .dll files, or other executables which are not located in the R software directory tree? No, it installs everything below R_HOME. We do rely on MCVCRT.DLL and some other system DLLs though, and I think this can get modified by other software. That's MSVCRT.DLL, the run-time library for MS Visual C++. It has been distributed with the OS since Win98 or so, and we don't touch it, so this is more of a possibility of other software interfering with us by replacing it with a bad version. 3. Does R modify the Windows registry in a non-obvious way, i.e. other than defining itself and what extensions to associate with R, and what are those extensions? I think all of its modifications would count as obvious. They are mainly below HKLM/Software/R-core or HKCU/Software/R-core (where the file locations are recorded); additionally file associations are set up for .Rdata files (which are called RWorkspace files there), and an uninstall entry is made. 4. Does R add macros to any part of MS Office? No. 5. Can you anticipate any other way in which installing and using R could disrupt the operation of another software? No, not really. Maybe users will become addicted to it? ;-) Duncan Murdoch *** Mat Soukup, Ph.D. Food and Drug Administration 10903 New Hampshire Ave. BLDG 22 RM 5329 Silver Spring, MD 20993-0002 Phone: 301.796.1005 *** ( I think Mat is owed a big Thank you for his efforts). Yes, indeed! Duncan Murdoch __ 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
[R] text mining with R
Hi all, Just wondering if anyone knows of any text mining projects in R...I googled a bit but didn't get anything... TIA, ken __ 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
Re: [R] quadratic form
On 3 Nov 2005, at 17:25, Robin Hankin wrote:
Hi Alvarez
If you define
quad.form.inv - function (M, x)
{
drop(crossprod(x, solve(M, x)))
}
then you will avoid an expensive call to %*% as well.
Is %*% really expensive in all platforms? I had a function that used QR
decomposition instead of quadratic forms, but then I got a message from
Canada suggesting that %*% would be faster. Indeed, it was in
not-too-large data sets and in Mac (or powerpc). I run some tests with
real applications, and found that my 800MHz iBook G4 run like a 2.5GHz
Intel machine when %*% was used. This really was architecture
dependent, since the performance boost was similar under OS X and Linux
in the very same PowerPC. So it seems that %*% is very cheap if you
have PowerPC, but it may be expensive in Intel. (I also run a test in
Sun, and it was somewhere between Intel and PowerPC.)
cheers, jari oksanen
--
Jari Oksanen, Oulu, Finland
__
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
Re: [R] quadratic form
Peter Dalgaard [EMAIL PROTECTED] writes:
Alvarez Pedro [EMAIL PROTECTED] writes:
On page 22 of the R-introduction guide it's written:
the quadratic form x^{'} A^{-1} x which is used in
multivariate computations, should be computed by
something like x%*%solve(A,x), rather than computing
the inverse of A.
Why isn't it good to compute t(x) %*% solve(A) %*% x?
It's just a waste of CPU time. For k x k matrices, solution of Ax=b is
of computational complexity O(k^2) whereas inversion of A is O(k^3).
This is obviously more important for k=1000 than for k=5.
Erm, Thomas Lumley points out that solution of Ax=b is also of order
k^3 ingeneral (R uses a QR decomposition). So it's just the multiplier
that differs. The savings should still be on the order of k^3 though.
--
O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B
c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K
(*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918
~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907
__
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
Re: [R] text mining with R
Just wondering if anyone knows of any text mining projects in R...I googled a bit but didn't get anything... RSiteSearch(text mining) turns up 85 hits... __ 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
Re: [R] fatal error unused tempdir
On 11/3/2005 11:12 AM, SANDRINE COELHO wrote: Hello, I am running R on XP Windows machine, and frequently I have enable to start R. I have this message: Fatal Error: cannot find unused tempdir name. I don't know why. This has come up before. When R starts up, it tries to create a folder for temporary files, using the TMP, TEMP or R_USER environment variables (in that order) to find a base path. The tempdir() function will tell you the folder name. It chooses the name by appending a random number between 0 and 65535 to a base name. It makes 100 attempts at this, and if it can't find a name that works, it bails out. That's what you're seeing. Sometimes it gets lucky and finds a name in the first 100 tries; then you don't get the error. R tries to delete the directory when it quits, but if there are files there, deletion will fail. So you should probably try to track down what is producing files and not getting rid of them. What you need to do is to find the directory containing all these temporary files (print tempdir() on one of your successful attempts), and delete all the old ones. Duncan Murdoch __ 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
[R] newbie graphics question: Two density plots in same frame ?
I swear I've scoured the help files and several texts before posting what feels like a dumb newbie question. How can I draw two kernel density plots in the same frame ? I have similar variables in two separate data frames, and I would like to show their two histograms/densities in a single picture. Same units, scale, range for both, so I'm simply trying to draw one and then add the other to the picture. Nothin' fancy. Bill Alpert Sr. Editor Barron's 212.416.2742 [EMAIL PROTECTED] [[alternative HTML version deleted]] __ 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
Re: [R] newbie graphics question: Two density plots in same frame ?
Here's a function that you can customize to fit your needs. lst is a named list.
multicomp - function(lst)
{
clr - c(darkgreen,red,blue,brown,magenta)
alldens - lapply(lst,function(x) {density(x,from=min(x),to=max(x))})
allx - sapply(alldens,function(d) {d$x})
ally - sapply(alldens,function(d) {d$y})
plot(allx,ally,type=n)
for (i in 1:length(lst)) {
lines(alldens[[i]]$x,alldens[[i]]$y,lty=i,col=clr[i],lwd=3)
}
legend(topright,xjust=1,legend=names(lst),lwd=3,lty=1:length(lst),col=head(clr,length(lst)))
}
#---
toplot - list(var1=dfr1$var,var2=dfr2$var)
multicomp(toplot)
-Original Message-
From: Alpert, William [mailto:[EMAIL PROTECTED]
Sent: Thursday, November 03, 2005 11:45 AM
To: r-help@stat.math.ethz.ch
Subject: [R] newbie graphics question: Two density plots in
same frame ?
I swear I've scoured the help files and several texts before posting
what feels like a dumb newbie question.
How can I draw two kernel density plots in the same frame ? I have
similar variables in two separate data frames, and I would
like to show
their two histograms/densities in a single picture. Same
units, scale,
range for both, so I'm simply trying to draw one and then add
the other
to the picture. Nothin' fancy.
Bill Alpert
Sr. Editor
Barron's
212.416.2742
[EMAIL PROTECTED]
[[alternative HTML version deleted]]
__
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
__
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
[R] Add dots at the mean of a bwplot using panel.points
How can I modify the example below to put a dot at the mean of each violin
plot? I assume I use panel.points but that's as far as I can go.
bwplot(voice.part ~ height, singer,
panel = function(..., box.ratio) {
panel.violin(..., col = transparent,
varwidth = FALSE, box.ratio = box.ratio)
#panel.points(mean(x.))
} )
TIA, Andy
__
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
Re: [R] newbie graphics question: Two density plots in same frame ?
To plot two Kernel densities you can use matplot: x1-density(rnorm(100)) x2-density(rnorm(100)) matplot(cbind(x1$y,x2$y), type=l) Or if both distributions are really very similar and you don't have to adjust the axes you can simply use plot(x1) lines(x2, col=red) Finally if you want to have two histograms in the same picture (I would not recomend it tough since the distributions are similar so the overlapping will make it very messy) you can use the argument add within hist hist(rnorm(100), col=red) hist(rnorm(100), col=blue, add=T) I hope this helps Francisco From: Alpert, William [EMAIL PROTECTED] To: r-help@stat.math.ethz.ch Subject: [R] newbie graphics question: Two density plots in same frame ? Date: Thu, 3 Nov 2005 11:45:21 -0500 I swear I've scoured the help files and several texts before posting what feels like a dumb newbie question. How can I draw two kernel density plots in the same frame ? I have similar variables in two separate data frames, and I would like to show their two histograms/densities in a single picture. Same units, scale, range for both, so I'm simply trying to draw one and then add the other to the picture. Nothin' fancy. Bill Alpert Sr. Editor Barron's 212.416.2742 [EMAIL PROTECTED] [[alternative HTML version deleted]] __ 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 __ 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
Re: [R] Visualizing a Data Distribution -- Was: breaks in hist()
Thanks for all the response. I think plotting a cdf or taking transformation could make the plot look better. But my further question is how to set the breaks to make the histogram concentrate in the interval of (0.01,0.2). I can even ignore the other parts of the values. Thanks! Leaf === At 2005-11-02, 12:07:12 you wrote: === Leaf Sun wrote: The histogram is highly screwed to the right, say, the range of the vector is [0, 2], but 95% of the value is squeezed in the interval (0.01, 0.2). I guess the histogram is as you wrote. See http://web.maths.unsw.edu.au/~tduong/seminars/intro2kde/ for a short explanation. -Original Message- From: Berton Gunter [mailto:[EMAIL PROTECTED] Sent: Wednesday, November 02, 2005 1:10 PM To: 'Leaf Sun'; r-help@stat.math.ethz.ch Subject: [R] Visualizing a Data Distribution -- Was: breaks in hist() Leaf: An interesting question concerning graphical perception. As you have noted, choice of bin boundaries in a histogram can have a big effect on how a distribution is perceived. My $.02 (U.S.): Histograms are a relic of manual data plotting. We have much better alternatives these days that should be used instead. e.g. 1. (my preference, but properly not consumer-friendly). Plot the cdf instead (?ecdf) . 2. Plot a density estimator (?density ; ?densityplot) 3. See David Scott's ash package, perhaps the KernSmooth package also (though density() probably already has anything that you'd need from it). Cheers, -- Bert Gunter Genentech Non-Clinical Statistics South San Francisco, CA The business of the statistician is to catalyze the scientific learning process. - George E. P. Box -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Leaf Sun Sent: Wednesday, November 02, 2005 9:49 AM To: r-help@stat.math.ethz.ch Subject: [R] breaks in hist() Dear listers, A quick question about breaks in hist(). The histogram is highly screwed to the right, say, the range of the vector is [0, 2], but 95% of the value is squeezed in the interval (0.01, 0.2). My question is : how to set the breaks then make the histogram look even? Thanks in advance, Leaf __ 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 = = = = = = = = = = = = = = = = = = = = __ 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
Re: [R] quadratic form
If you meant QR vs. inverting X'X for linear regression, the motivation for
using QR is not speed, but numerical stability. There's no univerally good
least squares algorithm that would be uniformly better than anything else
for any kind of data.
Andy
From: Jari Oksanen
On 3 Nov 2005, at 17:25, Robin Hankin wrote:
Hi Alvarez
If you define
quad.form.inv - function (M, x)
{
drop(crossprod(x, solve(M, x)))
}
then you will avoid an expensive call to %*% as well.
Is %*% really expensive in all platforms? I had a function
that used QR
decomposition instead of quadratic forms, but then I got a
message from
Canada suggesting that %*% would be faster. Indeed, it was in
not-too-large data sets and in Mac (or powerpc). I run some
tests with
real applications, and found that my 800MHz iBook G4 run like
a 2.5GHz
Intel machine when %*% was used. This really was architecture
dependent, since the performance boost was similar under OS X
and Linux
in the very same PowerPC. So it seems that %*% is very cheap if you
have PowerPC, but it may be expensive in Intel. (I also run a test in
Sun, and it was somewhere between Intel and PowerPC.)
cheers, jari oksanen
--
Jari Oksanen, Oulu, Finland
__
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
__
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
[R] nlme questions
Dear R users; Ive got two questions concerning nlme library 3.1-65 (running on R 2.2.0 / Win XP Pro). The first one is related to augPred function. Ive been working with a nonlinear mixed model with no problems so far. However, when the parameters of the model are specified in terms of some other covariates, say treatment (i.e. phi1~trt1+trt2, etc) the augPred function give me the following error: Error in predict.nlme(object, value[1:(nrow(value)/nL),,drop=FALSE], : Levels 0,1 not allowed for trt1, trt2. The same model specification as well as the augPred function under SPlus 2000 run without problems. The second question has to deal with the time needed for the model to converge. It really takes a lot of time to fit the model on R in relation to the time required to fit the same model on SPlus. I can imagine this is related to the optimization algorithm or something like that, but I would like to have a different opinion on these two issues. Thanks in advance Christian Mora __ 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
Re: [R] ML optimization question--unidimensional unfolding scaling
Hi Spencer Andy: Thanks for your thoughtful input! I did at one point look at the optim() function run debug on it (wasn't aware of browser--that's helpful!). My impression is that optim() simply calls a C function that handles the maximization. So if I want to break out of my likelihood function to restart optim() w/ new values, it seems I'd have to somehow communicate to C that it's time to stop. May need to rewrite the C, with which I'm not familiar--Java yes, so maybe when I have some real free time Another possibility might be finding some jerry-rigged way to break out of optim. Maybe if I tell the likelihood function to freeze its returned value at some point, optim will conclude it's done and stop. Probably inefficient I will have the problem of telling when the break point ought to occur. Just wish there were some programmatic way to say 'stop this and return control to the higher-level calling function 'blah''. A third possibility is one suggested by Spencer who seems to think it's ok for the routine to pursue multiple branches w/o restarting, hence no restart problem. But w/ Newtonian-style convergence the latent scale values (which are parameters to be estimated) have current positions are supposed to smoothly move toward lower likelihood values. What will happen in branched convergence, however, is that some of the latent values will prove to have better values on the other side of a normal curve from their current position. My guess is that this will cause the likelihood function to make a sudden, non-continuous jump not predictable by derivatives, which may mean it can't converge properly. Spencer's MDS alternative is intriguing I'll need to think more about it. Maybe I should also consider full-out Bayesian Monte Carlo methods (if I have time), which would simultaneously explore the whole solution space. Thanks, Peter On 11/2/05 9:01 PM, Spencer Graves [EMAIL PROTECTED] wrote: Have you looked at the code for optim? If you execute optim, it will list the code. You can copy that into a script file and walk through it line by line to figure out what it does. By doing this, you should be able to find a place in the iteration where you can test both branches of each bifurcation and pick one -- or keep a list of however many you want and follow them all more or less simultaneously, pruning the ones that seem too implausible. Then you can alternate between a piece of the optim code, bifurcating and pruning, adjusting each and printing intermediate progress reports to help you understand what it's doing and how you might want to modify it. With a bit more effort, you can get the official source code with comments. To do that, I think you go to www.r-project.org - CRAN - (select a local mirror) - Software: R sources. From there, just download The latest release: R-2.2.0.tar.gz. For more detailed help, I suggest you try to think of the simplest possible toy problem that still contains one of the issues you find most difficult. Then send that to this list. If readers can copy a few lines of R code from your email into R and try a couple of things in less than a minute, I think you might get more useful replies quicker. On 11/3/05 8:08 AM, Liaw, Andy [EMAIL PROTECTED] wrote: Alternatively, just type debug(optim) before using it, then step through it by hitting enter repeatedly... When you're done, do undebug(optim). On 11/3/05 11:06 AM, Liaw, Andy [EMAIL PROTECTED] wrote: Essentially all that debug() does is like inserting browser() as the first line of the function being debug()ed. You can type just about any command at the browser prompt, e.g., for checking data, etc. ?browser has list of special commands for the browser prompt. Andy __ 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
[R] Fitting heteroscedastic linear models/ problems with varIdent of nlme
Hi, I would like to fit a model for a factorial design that allows for unequal variances in all groups. If I am not mistaken, this can be done in lm by specifying weights. A function intended to specify weights for unequal variance structures is provided in the nlme library with the varIdent function. Is it apropriate to use these weights with lm? If not, is there another possibility to do factorial designs with heteroscedasticity? When trying to use varIdent I get an errormessage that says that varIndent is not a defined class. The function calls are written in the same way as in Pinheiro Bates Book. Their example works. With my data it doesnt. I somehow fail to figure out the difference between them. In the remainder is a subset of my Dataset in which the problem also occours. A,B,C are metric variables, D is a factor. I wold be thankful if any Ideas you might have. thank you for your attention Andreas dat2 A B C D 1 990.5 20 46 1 2 990.5 20 44 1 3 704.5 19 35 1 4 990.5 20 39 1 5 2240.5 25 79 2 6 2240.5 25 43 2 7 2240.5 25 44 2 8 2240.5 25 50 2 9 2240.5 25 56 2 10 470.0 17 51 2 vi-varIdent(form=~1|D) vi-initialize(vi,dat2) Error in getClass(Class) : c(\varIdent\ is not a defined class, \varFunc\ is not a defined class) In addition: Warning message: the condition has length 1 and only the first element will be used in: if (!is.na(match(Class, .BasicClasses))) return(newBasic(Class, __ 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
Re: [R] ML optimization question--unidimensional unfolding scalin g
Hi Spencer: Just realized I may have misunderstood your comments about branching--you may have been thinking about a restart. Sorry if I misrepresented them. See below: On 11/3/05 11:03 AM, Spencer Graves [EMAIL PROTECTED] wrote: Hi, Andy and Peter: That's interesting. I still like the idea of making my own local copy, because I can more easily add comments and test ideas while working through the code. I haven't used debug, but I think I should try it, because some things occur when running a function that don't occur when I walk through it line by line, e.g., parsing the call and ... arguments. Debug's handy tho I think it is line by line. Two more comments on the original question: 1. What is the structure of your data? Have you considered techniques for Multidimensional Scaling (MDS)? It seems that your problem is just a univariate analogue of the MDS problem. For metric MDS from a complete distance matrix, the solution is relatively straightforward computation of eigenvalues and vectors from a matrix computed from the distance matrix, and there is software widely available for the nonmetric MDS problem. For a terse introduction to that literature, see Venables and Ripley (2002) Modern Applied Statistics with S, 4th ed. (Springer, distance methods in sec. 11.1, pp. 306-308). I was looking for something on MDS in R, that'll be handy! The data structure is a set of variables (say about 6) that I have reason to believe measure an underlying dimension. I suspect that several of the variables are unfolding--that is, they have their highest value for some point on the scale and fall off w/ distance from that point in either direction. The degree of fall-off may vary depending on the variable. Some seem to fall off very rapidly, others not. A couple variables probably monotonically increase w/ the underlying scale, so they don't unfold. I can construct a distance matrix consisting of distances between these variables. Do you think MDS might be able to handle an arrangement like this, w/ some values folded about a scale point and with drop-off varying between variables? The distances between the variables do not map in any straightforward way into distances on the underlying scale because of folding and non-linearity. 2. If you don't have a complete distance matrix, might it be feasible to approach the problem starting small and building larger, i.e., start with 3 nodes, then add a fourth, etc.? Not sure I follow, but I do have a complete distance matrix of distances between the variables. spencer graves Thanks, Peter __ 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
[R] Specify Z matrix with lmer function
Is there a way to specify a Z matrix using the lmer function, where the model is written as y = X*Beta + Z*u + e? I am trying to reproduce smoothing methods illustrated in the paper Smoothing with Mixed Model Software my Long Ngo and M.P. Wand. published in the /Journal of Statistical Software/ in 2004 using the lme4 and Matrix packages. The code and data sets used can be found at http://www.jstatsoft.org/v09/i01/. There original code did not work for me without slight modifications here is the code that I used with my modifications noted. x - fossil$age y - 10*fossil$strontium.ratio knots - seq(94,121,length=25) n - length(x) X - cbind(rep(1,n),x) Z - outer(x,knots,-) Z - Z*(Z0) # I had to create the groupedData object with one group to fit the model I wanted grp - rep(1,n) grp.dat-groupedData(y~Z|grp) fit - lme(y~-1+X,random=pdIdent(~-1+Z),data=grp.dat) I would like to know how I could fit this same model using the lmer function. Specifically can I specify a Z matrix in the same way as I do above in lme? Thanks, Mark __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] multidimensional integration not over a multidimensionalrectangle
Hi,
anyone knows about any functions in R can get multidimensional integration
not over a multidimensional rectangle (not adapt).
For example, I tried the following function f(x,n)=x^n/n!
phi.fun-function(x,n)
{ if (n==1) {
x
}else{
integrate(phi.fun, lower=0, upper=x, n=n-1)$value
}
}
I could get f(4,2)=4^2/2!=8, but failed in f(4,3)=4^3/3! Thanks
Best,
Lynette
__
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
Re: [R] Fitting heteroscedastic linear models/ problems with varIdent of nlme
Andreas Cordes andreas.cordes at stud.uni-goettingen.de writes: Hi, I would like to fit a model for a factorial design that allows for unequal variances in all groups. If I am not mistaken, this can be done in lm by specifying weights. A function intended to specify weights for unequal variance structures is provided in the nlme library with the varIdent function. Is it apropriate to use these weights with lm? If not, is there another possibility to do factorial designs with heteroscedasticity? No, varIdent and friends is made for use with mixed effect models in package nlme. Dieter __ 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
[R] multivariate nonparametric regression with e = 0
Hello all, I'm a relatively new user of R, having mostly used it only for plotting so far. I'm also not very familiar with regression methods, hence forgive my greenness on the topic. What I want to do in R is multivariate nonparametric regression, with a slight hitch. From my experimental data I have a multitude of samples whose values approximate a function `f' that is defined over a 5D space (i.e., f: R^5-R). The values of the collected samples, call these `y', approximate `f', but due to the process by which they are collected, they always over-estimate (i.e., y = f + e, e = 0). The distribution of the error `e' can likely be modelled using the positive half of the normal distribution. Naturally I'm trying to obtain a smooth and relatively faithful approximation of `f' using the collected samples `y'. What would be the most fruitful approach in R to doing this? Even suggestions on which package/function to use would be tremendously helpful, as I don't yet know what their strengths/weaknesses are. Also, I would consider parametric regression as well, but in the general case I don't think I can assume/guess for my data at what the appropriate parametric basis functions should be... -- Maciej Kalisiak [EMAIL PROTECTED] http://www.dgp.toronto.edu/~mac/ __ 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
Re: [R] newbie graphics question: Two density plots in same frame ?
On 11/3/05, Alpert, William [EMAIL PROTECTED] wrote: I swear I've scoured the help files and several texts before posting what feels like a dumb newbie question. How can I draw two kernel density plots in the same frame ? I have similar variables in two separate data frames, and I would like to show their two histograms/densities in a single picture. Same units, scale, range for both, so I'm simply trying to draw one and then add the other to the picture. Nothin' fancy. Using densityplot from lattice: library(lattice) d1 = data.frame(x = rnorm(100)) d2 = data.frame(x = rnorm(100, mean = 0.5)) densityplot(~d1$x + d2$x, plot.points = FALSE) -Deepayan __ 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
Re: [R] Problems with pf() with certain noncentral values/degrees of freedom combinations
The problem is in src/nmath/pnbeta.c, which has an iteration limit of 100, not enough for these problems. Increasing the iteration limit to 1000 seems to work. -thomas On Mon, 24 Oct 2005, Ken Kelley wrote: Hello all. It seems that the pf() function when used with noncentral parameters can behave badly at times. I've included some examples below, but what is happening is that with some combinations of df and ncp parameters, regardless of how large the quantile gets, the same probability value is returned. Upon first glance noncentral values greater than 200 may seem large, but they are in some contexts not large at all. The problems with pf() can thus have serious implications (for example, in the context of sample size planning). I noticed that in in 1999 and 2000 issues with large degrees of freedom came about (PR#138), but I couldn't find the present issue reported anywhere. Might there be a way to make the algorithm more stable? I'm not sure how difficult this issue might be to fix, but hopefully it won't be too bad and can be easily done. Any thoughts on a workaround until then? Thanks, Ken Kelley # Begin example code X - seq(10, 600, 10) # Gets stuck at .99135 round(pf(X, 10, 1000, 225), 5) round(pf(X, 10, 200, 225), 5) round(pf(X, 5, 1000, 225), 5) round(pf(X, 5, 200, 225), 5) round(pf(X, 1, 1000, 225), 5) round(pf(X, 1, 200, 225), 5) # Gets stuck at .97035 round(pf(X, 10, 1000, 250), 5) round(pf(X, 10, 200, 250), 5) round(pf(X, 5, 1000, 250), 5) round(pf(X, 5, 200, 250), 5) round(pf(X, 1, 1000, 250), 5) round(pf(X, 1, 200, 250), 5) # Gets stuck at .93539 round(pf(X, 10, 1000, 275), 5) round(pf(X, 10, 200, 275), 5) round(pf(X, 5, 1000, 275), 5) round(pf(X, 5, 200, 275), 5) round(pf(X, 1, 1000, 275), 5) round(pf(X, 1, 200, 275), 5) # end example code version _ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major2 minor2.0 year 2005 month10 day 06 svn rev 35749 language R -- Ken Kelley, Ph.D. Inquiry Methodology Program Indiana University 201 North Rose Avenue, Room 4004 Bloomington, Indiana 47405 http://www.indiana.edu/~kenkel __ 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 Thomas Lumley Assoc. Professor, Biostatistics [EMAIL PROTECTED] University of Washington, Seattle __ 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
Re: [R] margins too large
Hi Sara Mouro wrote: Dear all, How can I explian and solve the error message: margins too large Do you mean figure margins too large? If so, it means that there is not enough room for your plot; try making the graphics window (or page size) bigger. Depending on how standard plot.fasp() is, you could also try reducing the plot margins by something like ... par(mar=rep(1, 4)) Paul which appears when I do something like: KK - alltypes(SpatData, K) plot.fasp(KK) Hope someone can please help me on this. Regards, Sara Mouro [[alternative HTML version deleted]] __ 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 -- Dr Paul Murrell Department of Statistics The University of Auckland Private Bag 92019 Auckland New Zealand 64 9 3737599 x85392 [EMAIL PROTECTED] http://www.stat.auckland.ac.nz/~paul/ __ 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
[R] Help in expand.grid() (Restricted combination)
Dear Rs: BY having the following code: candidates-expand.grid(e=c(nearest-neighbor,exaustive), d=c(70,75,80,85,90,92,94,96,98,99), n=c(20,25,30,35,40)) results in : e d n 1 nearest-neighbor 70 20 2 exaustive 70 20 3 nearest-neighbor 75 20 4 exaustive 75 20 90 exaustive 90 40 91 nearest-neighbor 92 40 92 exaustive 92 40 93 nearest-neighbor 94 40 94 exaustive 94 40 95 nearest-neighbor 96 40 96 exaustive 96 40 97 nearest-neighbor 98 40 98 exaustive 98 40 99 nearest-neighbor 99 40 100 exaustive 99 40 I need to associate nearest-neighbor with d=c(70,75,80,85,90,92,94,96,98,99), n=c(20,25,30,35,40) but exaustive only with d=c(70,75,80,85,90,92,94,96,98,99). Therefore I will have only 50+10 combinations not 100. I need to have combination as shown below 1 nearest-neighbor 70 20 2 exaustive 70 3 nearest-neighbor 75 20 4 exaustive 75 60 Thanks Prasanna -- Prasanna BALAPRAKASH IRIDIA, Université Libre de Bruxelles 50, Av. F. Roosevelt, CP 194/6 1050 Brussels Belgium. [[alternative HTML version deleted]] __ 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
Re: [R] RODBC and Excel: Wrong Data Type Assumed on Import
Kevin Wright [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED] From my experience (somewhat of a guess): Excel uses the first 16 rows of data to determine if a column is numeric or character. The data type which is most common in the first 16 rows will then be used for the whole column. I ran some experiments trying to force RODBC to read column 1 of my worksheet as character data (the data are mostly numbers with two exceptions, 275a and 275b, as mentioned earlier). Here's the base code: library(RODBC) channel - odbcConnectExcel(U:/efg/lab/R/Krumlauf-Plasmid/construct list.xls) plasmid - sqlFetch(channel,Sheet1, as.is=TRUE) odbcClose(channel) names(plasmid) [1] Plasmid Number PlasmidConcentration Comments Lost When Excel Sheet1 has rows 2:13 as an X to attempt to force treatment of column 1 as character data: class(plasmid$Plasmid Number) [1] numeric typeof(plasmid$Plasmid Number) [1] double plasmid$Plasmid Number[1:20] [1] NA NA NA NA NA NA NA NA NA NA NA NA 2 3 4 5 6 7 8 9 Why would any software with 12 consecutive X character strings assume the data are purely numeric? Add one more X so rows 2:14 have an X to attempt to force treatment of column 1 as character data: class(plasmid$Plasmid Number) [1] character typeof(plasmid$Plasmid Number) [1] character plasmid$Plasmid Number[1:20] [1] X X X X X X X X X X X X X NA NA NA NA NA NA NA So RODBC now recognizes character Xs in column 1 and then declares all numbers as invalid? These are incredibly (bad) assumptions. I say this is a bug, but it may be an ODBC problem and not one with R. And if this is not an official bug, then it's a serious design problem. Minimally, this issue should be described in the R Data Import/Export document, which everyone is told to read before asking a question. It's frustrating when packages like this work for toy problems, and the documentation never mentions the pitfalls of real data. The gregmisc bundle has a different read.xls function that uses a Perl script (xls2csv) and seems to be safer with mixed-type columns. Requires a working version of Perl. Thanks for this suggestion, but I think I'll just convert the Excel spreadsheet to a .csv and maintain it in that format. efg __ 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
Re: [R] Fitting heteroscedastic linear models/ problems with varIdent of nlme
Dieter Menne [EMAIL PROTECTED] writes: Andreas Cordes andreas.cordes at stud.uni-goettingen.de writes: Hi, I would like to fit a model for a factorial design that allows for unequal variances in all groups. If I am not mistaken, this can be done in lm by specifying weights. A function intended to specify weights for unequal variance structures is provided in the nlme library with the varIdent function. Is it apropriate to use these weights with lm? If not, is there another possibility to do factorial designs with heteroscedasticity? No, varIdent and friends is made for use with mixed effect models in package nlme. - or with generalized least squares, using the gls() function in the same package. -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ 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
Re: [R] Add dots at the mean of a bwplot using panel.points
How can I modify the example below to put a dot at the mean of each violin
plot? I assume I use panel.points but that's as far as I can go.
bwplot(voice.part ~ height, singer,
panel = function(..., box.ratio) {
panel.violin(..., col = transparent,
varwidth = FALSE, box.ratio = box.ratio)
#panel.points(mean(x.))
} )
Well, I answered my own question and learned something of the dark art of
lattice plots in the process. If this is an inane way to go about this, then
somebody please say so.
panel.mean - function(x,y,...){
y - as.numeric(y)
y.unique - sort(unique(y))
for(Y in y.unique) {
X - x[y == Y]
if (!length(X)) next
mean.value - list(x = mean(X), y = Y)
do.call(lpoints, c(mean.value, pch = 20))
}
}
bwplot(voice.part ~ height, singer,
panel = function(...) {
panel.violin(...,col = transparent)
panel.mean(...)
})
__
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
[R] Using R with SGE
Hello, I was wondering if there is a way to use R with the Sun Grid Engine, as opposed to using Rmpi. Our sge has a lammpi parrallel environment as well. Does anyone have a script to submit a batch R job using the SGE scheduler? Or can point me to a good tutorial? Thanks __ 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
[R] Error message: The following object(s) are masked
Hello!
First time posting here:
Here is my code:
x - c(1:22)
finaloutput=cidrm=NULL
finaldiversityoutput=diversitym=NULL
diversityinfo=read.table(Diversity_info.txt, header=T, sep=\t,
row.names=NULL)
attach(diversityinfo)
diversitynr=nrow(diversityinfo)
diversitytemp - matrix(0,nrow=diversitynr,ncol=1)
for(j in 1:length(x))
{
diversitym=read.table(paste(paste(Div-Chr,x[j],
sep=),_corr.txt,sep=), header=T, sep=\t, row.names=NULL)
attach(diversitym)
diversitytemp - cbind(diversitytemp,diversitym)
print(paste(paste(Diversity-Chrom,x[j], sep=),.txt,sep=))
print(dim(diversitytemp))
}
I am essentially trying to combine several tab-delimited text data files
together into one big file.
I recently upgraded to R 2.2.0. I now get multiple error messages of
the form:
The following object(s) are masked from diversitym ( position 4 ) :
X X.1 X.2
I searched on masked and read the manual about conflicts but I don't
really understand what the issue is. I didn't get this error message,
using the same code with R 2.1.0. Can somebody help (I'm not a
programmer).
[[alternative HTML version deleted]]
__
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
Re: [R] Help in expand.grid() (Restricted combination)
If I understand you correctly, you cannot do this with a data.frame, which must be rectangular with equal numbers of entries (columns) in each row. See ?.data.frame and read An Introduction to R for these basics. You could make the 3rd column for exhaustive = NA (or maybe an empty string, ) I suppose, and rbind the results you want: part1-expand.grid(e='nearest-neighbor',d=c(70,75,80,85,90,92,94,96,98,99), n=c(20,25,30,35,40)) part2-data.frame(e='exhaustive',d=c(70,75,80,85,90,92,94,96,98,99),n=NA) result-rbind(part1,part2) -- Bert Gunter Genentech Non-Clinical Statistics South San Francisco, CA The business of the statistician is to catalyze the scientific learning process. - George E. P. Box -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Prasanna Sent: Thursday, November 03, 2005 12:07 PM To: r-help@stat.math.ethz.ch Subject: [R] Help in expand.grid() (Restricted combination) Dear Rs: BY having the following code: candidates-expand.grid(e=c(nearest-neighbor,exaustive), d=c(70,75,80,85,90,92,94,96,98,99), n=c(20,25,30,35,40)) results in : e d n 1 nearest-neighbor 70 20 2 exaustive 70 20 3 nearest-neighbor 75 20 4 exaustive 75 20 90 exaustive 90 40 91 nearest-neighbor 92 40 92 exaustive 92 40 93 nearest-neighbor 94 40 94 exaustive 94 40 95 nearest-neighbor 96 40 96 exaustive 96 40 97 nearest-neighbor 98 40 98 exaustive 98 40 99 nearest-neighbor 99 40 100 exaustive 99 40 I need to associate nearest-neighbor with d=c(70,75,80,85,90,92,94,96,98,99), n=c(20,25,30,35,40) but exaustive only with d=c(70,75,80,85,90,92,94,96,98,99). Therefore I will have only 50+10 combinations not 100. I need to have combination as shown below 1 nearest-neighbor 70 20 2 exaustive 70 3 nearest-neighbor 75 20 4 exaustive 75 60 Thanks Prasanna -- Prasanna BALAPRAKASH IRIDIA, Universiti Libre de Bruxelles 50, Av. F. Roosevelt, CP 194/6 1050 Brussels Belgium. [[alternative HTML version deleted]] __ 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
Re: [R] Potential for R to conflict with other softwares
I just wanted to make one clarification about my statement: After some time, my collegues at the Food and Drug Adminstration have finally acknowledged R as a powerful statistical computing environment. I did not intend for this to read that R has been acknowledged as being 21 CFR Part 11 compliant. This is a whole other ball game. What I meant to say is that, within the Center of Drug Evaluation and Research (CDER) the Office of Biostatistics is willing to look into whether or not reviewers can download R onto their government issued PC's. And this must all be approved by the Office of Information and Technology. So admittedly, this is only a small step, but nonetheless it is a step in the right direction. I apologize for any confusion. Cheers, Mat Disclaimer which I also forgot in the original post: The following views are those of the writer and do not necessarily reflect those of the FDA. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Duncan Murdoch Sent: Thursday, November 03, 2005 6:30 AM To: Soukup, Mat Cc: 'r-help@stat.math.ethz.ch' Subject: Re: [R] Potential for R to conflict with other softwares On 11/3/2005 9:11 AM, Soukup, Mat wrote: Hi. After some time, my collegues at the Food and Drug Adminstration have finally acknowledged R as a powerful statistical computing environment. However, in order to comply with the Office of Information and Technology standards there are a couple of questions about whether R could interfere with other software. As I'm more of a driver of the R software and not a mechanic, I was hoping for the insight of the many great useRs. Below is a list of 5 proposed questions to which I value any comment. Thank you for your time, -Mat These answers are about the Windows version only, but from the questions, I think that's what you were looking for. They apply to all versions since 1.6.x at least (though the earlier ones would have put fewer entries into the registry, they put them in the same places). 1. Does R have high resolution graphics? Yes, but I don't think I get the point of this question. How would that interfere with other software? 2. Does R have .dll files, or other executables which are not located in the R software directory tree? No, it installs everything below R_HOME. 3. Does R modify the Windows registry in a non-obvious way, i.e. other than defining itself and what extensions to associate with R, and what are those extensions? I think all of its modifications would count as obvious. They are mainly below HKLM/Software/R-core or HKCU/Software/R-core (where the file locations are recorded); additionally file associations are set up for .Rdata files (which are called RWorkspace files there), and an uninstall entry is made. 4. Does R add macros to any part of MS Office? No. 5. Can you anticipate any other way in which installing and using R could disrupt the operation of another software? No, not really. Maybe users will become addicted to it? ;-) Duncan Murdoch ** * Mat Soukup, Ph.D. Food and Drug Administration 10903 New Hampshire Ave. BLDG 22 RM 5329 Silver Spring, MD 20993-0002 Phone: 301.796.1005 ** * [[alternative HTML version deleted]] __ 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 __ 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 __ 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
Re: [R] Add dots at the mean of a bwplot using panel.points
On 11/3/05, Andy Bunn [EMAIL PROTECTED] wrote:
How can I modify the example below to put a dot at the mean of each
violin
plot? I assume I use panel.points but that's as far as I can go.
bwplot(voice.part ~ height, singer,
panel = function(..., box.ratio) {
panel.violin(..., col = transparent,
varwidth = FALSE, box.ratio = box.ratio)
#panel.points(mean(x.))
} )
Well, I answered my own question and learned something of the dark art of
lattice plots in the process. If this is an inane way to go about this,
then
somebody please say so.
panel.mean - function(x, y, ...) {
tmp - tapply(x, y, FUN = mean)
panel.points(tmp, seq(tmp), pch = 20, ...)
}
is more direct, but otherwise your solution is fine.
-Deepayan
panel.mean - function(x,y,...){
y - as.numeric(y)
y.unique - sort(unique(y))
for(Y in y.unique) {
X - x[y == Y]
if (!length(X)) next
mean.value - list(x = mean(X), y = Y)
do.call(lpoints, c(mean.value, pch = 20))
}
}
bwplot(voice.part ~ height, singer,
panel = function(...) {
panel.violin(...,col = transparent)
panel.mean(...)
})
__
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
Re: [R] RODBC and Excel: Wrong Data Type Assumed on Import
You could try using the COM interface rather than the ODBC interface. Try code such as this: library(RDCOMClient) xls - COMCreate(Excel.Application) xls[[Workbooks]]$Open(MySpreadsheet.xls) sheet - xls[[ActiveSheet]] mydata - sheet[[UsedRange]][[value]] xls$Quit() # convert mydata to a character matrix mydata.char - matrix(unlist(mydata), nc = length(xx)) On 11/3/05, Kevin Wright [EMAIL PROTECTED] wrote: From my experience (somewhat of a guess): 1. Excel uses the first 16 rows of data to determine if a column is numeric or character. The data type which is most common in the first 16 rows will then be used for the whole column. If you sort the data so that at least the first 9 rows have character data, you may find this allows the data to be interpreted as character. There is supposedly a registy setting that can control how many lines to use (instead of 16), but I have not had success with the setting. I suspect that ODBC uses JET4, which may be the real source of the problem. See more here: http://www.dicks-blog.com/archives/2004/06/03/external-data-mixed-data-types/ 2. The gregmisc bundle has a different read.xls function that uses a Perl script (xls2csv) and seems to be safer with mixed-type columns. Requires a working version of Perl. Best, Kevin Wright The first column in my Excel sheet has mostly numbers but I need to treat it as character data: library(RODBC) http://tolstoy.newcastle.edu.au/R/help/05/09/11324.html#14938qlink1 * channel - odbcConnectExcel(U:/efg/lab/R/Plasmid/construct list.xls) * * plasmid - sqlFetch(channel,Sheet1, as.is=TRUE) * * odbcClose(channel) * names(plasmid) [1] Plasmid Number Plasmid Concentration Comments Lost # How is the type decided? I need a character type. class(plasmid$Plasmid Number) [1] numeric typeof(plasmid$Plasmid Number) [1] double plasmid$Plasmid Number[273:276] [1] 274 NA NA 276 The two NAs are supposed to be 275a and 275b. I tried the as.is=TRUE but that didn't help. I consulted Section 4, Relational databases, in the R Data Import/Export document (for Version 2.2.0). Section 4.2.2, Data types, was not helpful. In particular, this did not seem helpful: The more comprehensive of the R interface packages hide the type conversion issues from the user. Section 4.3.2, Package RODBC, provided a simple example of using ODBC .. with a(sic) Excel spreadsheet but is silent on how to control the data type on import. Could the documentation be expanded to address this issue? I really need to show Plasmid 275a and Plasmid 275b instead of Plasmid NA. Thanks for any help with this. efg -- Earl F. Glynn Scientific Programmer Bioinformatics Department [[alternative HTML version deleted]] __ 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 __ 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
Re: [R] FIGARCH
RSiteSearch(FIGARCH) revealed that Diethelm Wuertz prepared an R
interface to Ox Garch. However, Ox and all its components are
copyright of Jurgen A. Doornik. The Console (command line) versions may
be used freely for academic research and teaching purposes only.
Commercial users and others who do not qualify for the free version must
purchase the Windows version of Ox ... . See
http://finzi.psych.upenn.edu/R/library/fSeries/html/A3-GarchOxModelling.html;.
I don't know if this will help you.
Best Wishes,
Spencer Graves
Sumanta Basak wrote:
Hi All,
Currently I'm working in FIGARCH process [Fractionally Integrated
Generalized Autoregressive Conditional Heteroscedasticity]. I've already
got the codes to do the process in S-Plus. Can anyone help me to do it
in R?
Thanks,
SUMANTA BASAK.
---
This e-mail may contain confidential and/or privileged infor...{{dropped}}
__
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
--
Spencer Graves, PhD
Senior Development Engineer
PDF Solutions, Inc.
333 West San Carlos Street Suite 700
San Jose, CA 95110, USA
[EMAIL PROTECTED]
www.pdf.com http://www.pdf.com
Tel: 408-938-4420
Fax: 408-280-7915
__
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
[R] Plotting Factorial GLMs
Hello all, I'm attempting to plot the functions from a generalized linear model while iterating over multiple levels of a factor in the model. In other words, I have a data set Block, Treatment.Level, Response.Level So, the glm and code to plot should be logit.reg-glm(formula = Response.Level ~ Treatment.Level + Block, family=quasibinomial(link=logit))) plot( Response.Level ~ Treatment.Level) logit.reg.function - function (trt, blk) predict(logit.reg, data.frame(Treatment.Level=trt, Block=blk) curve(logit.reg.function(x, A), add=TRUE) But I get the error: Error in xy.coords(x, y) : 'x' and 'y' lengths differ Now, if I set Block=A in the function, and take blk out, as well as taking the A out of the curve statement, it plots just fine. What am I doing wrong, as this would be a nice, quick, and easy way to whip up multiple curves from a factorial dataset! __ 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
Re: [R] Search within a file
On 3 Nov 2005, [EMAIL PROTECTED] wrote: I am looking for a way to search a file for position of some expression, from within R. My current code: sha1Pos = gregexpr(sha1, readChar(filename, file.info(filename)$size))[[1]] Works fine for small files, but text files I will be working with might get up to Gb range, so I was trying to accomplish the same without loading the whole file into R. I would think you could use readLines to read in a batch of lines, run (g)regexpr, and keep track of matches and position. Create a connection to the file using file() first, and then subsequent calls to readLines will start where you left off. But you will need to adjust the position indices returned by gregexpr by how far into the file you are. Seems very doable. + seth __ 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
Re: [R] Problems with abline adding regression line to a graph
Hi
On 3 Nov 2005 at 16:03, CG Pettersson wrote:
Date sent: Thu, 03 Nov 2005 16:03:05 +0100
From: CG Pettersson [EMAIL PROTECTED]
To: r-help@stat.math.ethz.ch
Subject:[R] Problems with abline adding regression line to a
graph
Hello all,
R2.1.1, W2k
I try to make a plot of a simple regression model in this way:
with(njfA_bcd, {
+ plot(TC_OS.G31,Prot,cex = 2, col = red, xlab= TC/OS at GS32,
+ ylab=Grain crude protein (CP))
+ })
This part works well and produces the datapoints as red circles.
When I try to add a line, using a fitted linear model in a way
that works perfect with other variables in the same dateset the
following happens:
with(njfA_bcd, {
+abline(lm(predict(m1tc) ~ TC_OS.G31), lty = 1, col = red)
+ })
Error in model.frame(formula, rownames, variables, varnames, extras,
extranames, :
variable lengths differ
And this means?
There exists missing values for TC_OS.G31 in the dataset. From the
beginning m1tc was a lm() object, which gave the same Error message.
To try to fix the problem I changed to lme() and used
na.action=na.omit explicitely, but this didn´t help.
na.action=na.exclude
should help.
It should be probably mentioned in lm help page, at least as a
hyperlink to na.* parametrs.
HTH
Petr
Here is the summary of m1tc:
summary(m1tc)
Linear mixed-effects model fit by REML
Data: njfA_bcd
AIC BIClogLik
209.4914 219.0692 -100.7457
Random effects:
Formula: ~1 | Trial
(Intercept) Residual
StdDev:1.242184 0.6520464
Fixed effects: Prot ~ TC_OS.G31
Value Std.Error DF t-value p-value
(Intercept) 14.86209 0.957630 68 15.519662 0
TC_OS.G31 -24.22286 4.792801 68 -5.054008 0
Correlation:
(Intr)
TC_OS.G31 -0.935
Standardized Within-Group Residuals:
Min Q1 Med Q3 Max
-1.68329774 -0.73751040 -0.05600477 0.68301243 2.21693174
Number of Observations: 83
Number of Groups: 14
What is happening and what shall I do about it?
Cheers
/CG
--
CG Pettersson, MSci, PhD Stud.
Swedish University of Agricultural Sciences (SLU)
Dept. of Crop Production Ecology. Box 7043.
SE-750 07 UPPSALA, Sweden.
+46 18 671428, +46 70 3306685
[EMAIL PROTECTED]
__
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
Petr Pikal
[EMAIL PROTECTED]
__
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
