[R] Starting R within an VBA makro
Hello everybody, at work I want to start R within an VBA program. First I calculate something in Excel 2003. After that I want to start within my VBA makro R wich should start an R file, say i.g, superplot.R which plots me the data well. So is it possible? Maybe something like that: sub test() 'calculate something start R superplot.R end test() Thank you so much. Kindly regards, Maja -- Pt! Schon vom neuen GMX MultiMessenger gehört? Der kanns mit allen: http://www.gmx.net/de/go/multimessenger __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Tools For Preparing Data For Analysis
(Ted Harding) sent the following at 10/06/2007 09:28: ... much snipped ... (As is implicit in many comments in Robert's blog, and indeed also from many postings to this list over time and undoubtedly well known to many of us in practice, a lot of the problems with data files arise at the data gathering and entry stages, where people can behave as if stuffing unpaired socks and unattributed underwear randomly into a drawer, and then banging it shut). And they look surprised when pointing a statistician at the chest of drawers doesn't result in a cut price display worthy of Figleaf (or Victoria's Secret I think for those of you in N.America) and get them their degree, doctorate, latest publication ... Ah me, how wonderfully, wonderfully ... sadly, accurate! Thanks Ted, great thread and I'm impressed with EpiData that I've discovered through this. I'd still like something that is even more integrated with R but maybe some day, if EpiData go fully open source as I think they are doing (A full conversion plan to secure this and convert the software to open-source has been made (See complete description of license and principles). at http://www.epidata.dk/ but the link to http://www.epidata.dk/about.htm doesn't exactly clarify this I don't think. But I can hope.) Thanks, yet again, to everyone who creates and contributes to the R system and this list: wonderful! C -- Chris Evans [EMAIL PROTECTED] Skype: chris-psyctc Professor of Psychotherapy, Nottingham University; Consultant Psychiatrist in Psychotherapy, Notts PDD network; Research Programmes Director, Nottinghamshire NHS Trust; *If I am writing from one of those roles, it will be clear. Otherwise* *my views are my own and not representative of those institutions* __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] GMM estimation
Dear everyone: I have to finish my thesis to graduate as Bs. in Economics. I choose to estimate a New Keynesian Phillips Curve (NKPC) for Uruguay using Generalized Moment Method (GMM). I do not know programming or R but I would like to use it. Should I use gee, geepack or gam? Thanks in advance, Sebastián. *** ¡Hola todos! Para terminiar mi licenciatura en Economía debo hacer un trabajo de investigación monográfico. Elegi como tema la estimación de la curva de Phillips de los Nuevos Keynesianos (CPNK). No se programar ni conosco el lenguaje R pero me gustaria usarlo para estimar la CPNK usando el método generalizado de los momentos (MGM). ¿Debería usar el paquete gee, geepack o gam? Gracias a todos. Sebastián. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] generalized moment method
Dear everyone: I have to finish my thesis to graduate as Bs. in Economics. I choose to estimate a New Keynesian Phillips Curve (NKPC) for Uruguay using Generalized Moment Method (GMM). I do not know programming or R but I would like to use it. Should I use gee, geepack or gam? Thanks in advance, Sebastián. *** ¡Hola todos! Para terminiar mi licenciatura en Economía debo hacer un trabajo de investigación monográfico. Elegi como tema la estimación de la curva de Phillips de los Nuevos Keynesianos (CPNK). No se programar ni conosco el lenguaje R pero me gustaria usarlo para estimar la CPNK usando el método generalizado de los momentos (MGM). ¿Debería usar el paquete gee, geepack o gam? Gracias a todos. Sebastián. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Textpad help
I have installed textpad and tried running R code. But it gives me the following error message. The filename, directory name, or volume label syntax is incorrect. Tool completed with exit code 1 Can you provide any help? I am not a technical person. So the help in detail will be appreciated. Thx [[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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] icc from GLMM?
Shinichi Nakagawa S.Nakagawa at sheffield.ac.uk writes:
...
I am a little confused which one to trust and use. Or there are no easy form
to do this? I am guessing formula would change depending on what distribution
you use and what link function as well? I want to calculate icc from GLMM with
Poisson with log link function and also binomial with logit function. Could
anybody help me please?
Yes, you are right that ICC depends on assumed data distribution. While ICC is
very handy in linear models it is not the case in GLMM. I suggest you take a
look at the references bellow. There is also some online material by the same
authors. Additionally, I remember that there were lively discussions about ICC
on multilevel list at
http://www.jiscmail.ac.uk/lists/multilevel.html
Best wishes, Gregor
@Article{Goldstein:2002,
author = {Goldstein, H. and Browne, W. and Rabash, J.},
title ={Partitioning variation in multilevel models},
journal = {Understanding Statistics},
year = {2002},
volume = {1},
number = {4},
pages ={223--231},
keywords = {variance ratio, variance partition coefficient,
intra-unit correlation, intra-class correlation, normal
models, discrete models, random coefficient models}
}
@Article{Browne:2005,
author = {Browne, W. J. and Subramanian, S. V. and Jones, K. and
Goldstein, H.},
title ={Variance partitioning in multilevel logistic models that
exhibit overdispersion},
journal = {J. R. Stat. Soc. A Stat. Soc.},
year = {2005},
volume = {168},
number = {3},
pages ={599--613},
doi = {10./j.1467-985X.2004.00365.x},
checked = {[2006-04-16]},
keywords = {heritability, ratios, intra-class correlation,
intra-unit correlation, simulation, linearization, latent
variable approach},
}
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Starting R within an VBA makro
On Mon, 11 Jun 2007, Maja Schröter wrote: Hello everybody, at work I want to start R within an VBA program. First I calculate something in Excel 2003. After that I want to start within my VBA makro R wich should start an R file, say i.g, superplot.R which plots me the data well. So is it possible? http://cran.r-project.org/contrib/extra/dcom/00ReadMe.html and RSiteSearch(Excel VBA) Maybe something like that: sub test() 'calculate something start R superplot.R end test() Thank you so much. Kindly regards, Maja -- 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 and provide commented, minimal, self-contained, reproducible code.
[R] epitools and R 2.5
At work after updating to R 2.5 I get an error using epitab from package epitools, when at home (R 2.4) I get no error. Could someone help me? Thanks Pietro Bulian Servizio di Onco-Ematologia Clinico-Sperimentale I.R.C.C.S. Centro di Riferimento Oncologico Via Franco Gallini 2 33081 Aviano (PN) - Italy phone: +39 0434 659 412 fax: +39 0434 659 409 e-mail: [EMAIL PROTECTED] (at work) epitab(matrix(c(227,473,74,126),2)) the part of the args list of 'list' being evaluated was: (tab = tab, measure = oddsratio, conf.level = conf.level, pvalue = pvalue, ) Error in epitab(matrix(c(227, 473, 74, 126), 2)) : element 5 is empty R.version _ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 5.0 year 2007 month 04 day23 svn rev41293 language R version.string R version 2.5.0 (2007-04-23) Package: epitools Version: 0.4-8 Date: 2007-05-10 Title: Epidemiology Tools Author: Tomas Aragon [EMAIL PROTECTED] Maintainer: Tomas Aragon [EMAIL PROTECTED] Depends: R (= 2.1.0) Description: A package of analytic tools for epidemiologists. License: GPL version 2 or newer URL: http://www.epitools.net Packaged: Thu May 10 01:55:28 2007; Tomas Built: R 2.5.0; ; 2007-05-10 14:22:09; windows (at home) epitab(matrix(c(227,473,74,126),2)) Warning: a final empty element has been omitted the part of the args list of 'list' being evaluated was: (tab = tab, measure = oddsratio, conf.level = conf.level, pvalue = pvalue, ) $tab Outcome Predictor Disease1 p0 Disease2 p1 oddsratio lower upper p.value Exposed1 227 0.3242857 74 0.37 1.000 NA NA NA Exposed2 473 0.6757143 126 0.63 0.8171533 0.5887731 1.134120 0.2348794 $measure [1] wald $conf.level [1] 0.95 $pvalue [1] fisher.exact R.version _ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 4.1 year 2006 month 12 day 18 svn rev 40228 language R Package: epitools Title: Epidemiology Tools Version: 0.4-7 Date: 2005-03-20 Author: Tomas Aragon Description: Basic tools for applied epidemiology. Maintainer: Tomas Aragon [EMAIL PROTECTED] License: GPL version 2 or newer URL: http://www.epitools.net Packaged: Sun Mar 20 16:30:07 2005; Tomas Built: R 2.4.0; ; 2006-10-03 22:02:05; windows version.string R version 2.4.1 (2006-12-18) __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] epitools and R 2.5
On Mon, 11 Jun 2007, Pietro Bulian wrote: At work after updating to R 2.5 I get an error using epitab from package epitools, when at home (R 2.4) I get no error. Could someone help me? The maintainer: this is a long-standing bug in the package. But you have enough information from the error message to correct the bug and rebuild the package yourself. There are no such versions of R as '2.5' and '2.4' (see the posting guide), but R 2.4.0 did give a warning on your example. Note that you are using different versions of epitools in your two locations, a difference you failed to mention and which may be important. Thanks Pietro Bulian Servizio di Onco-Ematologia Clinico-Sperimentale I.R.C.C.S. Centro di Riferimento Oncologico Via Franco Gallini 2 33081 Aviano (PN) - Italy phone: +39 0434 659 412 fax: +39 0434 659 409 e-mail: [EMAIL PROTECTED] (at work) epitab(matrix(c(227,473,74,126),2)) the part of the args list of 'list' being evaluated was: (tab = tab, measure = oddsratio, conf.level = conf.level, pvalue = pvalue, ) Error in epitab(matrix(c(227, 473, 74, 126), 2)) : element 5 is empty R.version _ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 5.0 year 2007 month 04 day23 svn rev41293 language R version.string R version 2.5.0 (2007-04-23) Package: epitools Version: 0.4-8 Date: 2007-05-10 Title: Epidemiology Tools Author: Tomas Aragon [EMAIL PROTECTED] Maintainer: Tomas Aragon [EMAIL PROTECTED] Depends: R (= 2.1.0) Description: A package of analytic tools for epidemiologists. License: GPL version 2 or newer URL: http://www.epitools.net Packaged: Thu May 10 01:55:28 2007; Tomas Built: R 2.5.0; ; 2007-05-10 14:22:09; windows (at home) epitab(matrix(c(227,473,74,126),2)) Warning: a final empty element has been omitted the part of the args list of 'list' being evaluated was: (tab = tab, measure = oddsratio, conf.level = conf.level, pvalue = pvalue, ) $tab Outcome Predictor Disease1 p0 Disease2 p1 oddsratio lower upper p.value Exposed1 227 0.3242857 74 0.37 1.000 NA NA NA Exposed2 473 0.6757143 126 0.63 0.8171533 0.5887731 1.134120 0.2348794 $measure [1] wald $conf.level [1] 0.95 $pvalue [1] fisher.exact R.version _ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 4.1 year 2006 month 12 day 18 svn rev 40228 language R Package: epitools Title: Epidemiology Tools Version: 0.4-7 Date: 2005-03-20 Author: Tomas Aragon Description: Basic tools for applied epidemiology. Maintainer: Tomas Aragon [EMAIL PROTECTED] License: GPL version 2 or newer URL: http://www.epitools.net Packaged: Sun Mar 20 16:30:07 2005; Tomas Built: R 2.4.0; ; 2006-10-03 22:02:05; windows version.string R version 2.4.1 (2006-12-18) __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- 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 and provide commented, minimal, self-contained, reproducible code.
[R] Looking for R-code for non-negative matrix factorization in the presence of Gaussian or Poisson noise
Hi all, Has any of you implemented code for non-negative matrix factorization to solve Y=T P' +E; dim(Y)=n,p ; dim(T)=n,nc; dim (P)=(p,nc); dim(E)=n,p where T and P must be non-negative and E either Gaussian or Poisson noise. I'm looking for two variants: 1. Easy (I think), T is known (that is we just want to solve the general inverse problem) 2. Harder (?), T is unknown (under some restrictions) [as an intermediate, we may want to fix nc] Solutions/suggestions welcome Have a nice day, Christian Ritter Senior Statistical Consultant Analytical and Statistical Services Shell Coordination Centre S.A. Monnet Centre International Laboratory, Avenue Jean Monnet 1, B-1348 Louvain-La-Neuve, Belgium Tel: +32 10 477 349 Fax: +32 10 477 219 Email: [EMAIL PROTECTED] Internet: http://www.analyticalsolutions4u.eu [[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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] Lines in dotchart dotplot ?
--- [EMAIL PROTECTED] wrote: On 6/9/07, John Kane [EMAIL PROTECTED] wrote: Is it possible to use dotchart or dotplot and set the lines in such a way that they only extend from the left y-axis to the data point? Yes (sort of) in dotplot at least. E.g., dotplot(VADeaths, groups = FALSE, type = c(p, h)) dotplot(VADeaths, groups = FALSE, type = c(p, h), origin = 0) -Deepayan Ah, that is quite nice, not exactly what I remember from Cleveland but it should do quite nicely. Thank you very much I seem to remember that Wm Cleveland did this in his 1985 book The elements of graphing data. In cases where one has a true starting or O point on the x-scale this layout seems to be very effective in displaying some data. I know that I can do it by simple ploting lines and points but a more polished function than I am likely to produce would be nice. 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 and provide commented, minimal, self-contained, reproducible code.
[R] system() and R BATCH
If I start from within R a new R batch job by using something like system(R CMD BATCH --no-save --quiet Rin.txt Rout.txt, intern=FALSE, ignore.stderr=TRUE, wait=FALSE, input=NULL) the job runs fine and smooth. However, when, for any reason, I put twice ctrl+C in the calling R, it kills me the called batch job, too. This is not what I wanted. Ctrl +C is usually rather used to stop a loop etc. Why it stops the called batch jobs, too? What can be the solution for me - avoiding the called batch jobs to be killable from the calling R process? Best regards Meinhard PS system: MacOS 10.4.9 Intel, R 2.5.0 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Textpad help
Hi Mary, You really have not given us much to go on. An example of the code that you were trying to run would be a great help (as it says down below : PLEASE do read the posting guide. The point about code is important. However the first thing to check is your file path. This is my first guess anyway. Here is the entry in the R FAQ As R uses C-style string handling, `\' is treated as an escape character, so that for example one can enter a newline as `\n'. When you really need a `\', you have to escape it with another `\'. Thus, in filenames use something like c:\\data\\money.dat. You can also replace `\' by `/' (c:/data/money.dat). -- I am not familiar with textpad but you might want to have a look at tinn-r which is very closely integrated with R as an alternative editor --- Mary Royerr [EMAIL PROTECTED] wrote: I have installed textpad and tried running R code. But it gives me the following error message. The filename, directory name, or volume label syntax is incorrect. Tool completed with exit code 1 Can you provide any help? I am not a technical person. So the help in detail will be appreciated. Thx [[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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] generalized moment method
Sebastian Kruk wrote: Dear everyone: I have to finish my thesis to graduate as Bs. in Economics. I choose to estimate a New Keynesian Phillips Curve (NKPC) for Uruguay using Generalized Moment Method (GMM). I do not know programming or R but I would like to use it. Should I use gee, geepack or gam? Dear Sebastián -- neither geepack nor gam provide GMM estimators. GMM -- or at least minimum distance estimation techniques -- rely on fitting by linear or more often non-linear least squares functions of smaller parameter vectors to the empirical moments of your problem. R is a suitable tool for this, but there is AFAIK know general GMM package. The details of our model would need to be known before any further advice can be given. Regards, Markus Thanks in advance, Sebastián. *** ¡Hola todos! Para terminiar mi licenciatura en Economía debo hacer un trabajo de investigación monográfico. Elegi como tema la estimación de la curva de Phillips de los Nuevos Keynesianos (CPNK). No se programar ni conosco el lenguaje R pero me gustaria usarlo para estimar la CPNK usando el método generalizado de los momentos (MGM). ¿Debería usar el paquete gee, geepack o gam? Gracias a todos. Sebastián. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Markus Jantti Abo Akademi University [EMAIL PROTECTED] http://www.iki.fi/~mjantti __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] problem with xlsReadWrite package
Hi friends, I have installed R 2.4.0 in my pc. I have a file xls entitled dali following this directory:c://programfiles//R 2.4.0. Recently I have installed xlsreadwrite 1.3.2. but , when I wrote the following lines: library(xlsReadWrite) read.xls( file, colNames = TRUE, sheet = 1, type = data.frame, from = 1, colClasses = NA ) I obtained from R console the following messages: Error in library(xlsReadWrite) impossible to find the functionread.xls. Please help me, many thanks in advance. [[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 and provide commented, minimal, self-contained, reproducible code.
[R] problem with xlsReadWrite package
Hi friends, I have installed R 2.4.0 in my pc. I have a file xls entitled dali following this directory:c://programfiles//R 2.4.0. Recently I have installed xlsreadwrite 1.3.2. but , when I wrote the following lines: library(xlsReadWrite) read.xls( file, colNames = TRUE, sheet = 1, type = data.frame, from = 1, colClasses = NA ) I obtained from R console the following messages: Error in library(xlsReadWrite) impossible to find the functionread.xls. Please help me, many thanks in advance. [[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 and provide commented, minimal, self-contained, reproducible code.
[R] Error using mgcv package
Hi all, I need some solution in the following problem. The following error appears when i use mgcv package for implementing GAM. But the same formula works fine in gam package. model.gam - gam(formula = RES ~ CAT01+s(NUM01,5)+CAT02+CAT03+s(NUM02,5)+CAT04+ + CAT05+s(NUM03,5)+CAT06+CAT07+s(NUM04,5)+CAT08+s(NUM05,5)+CAT09+ + CAT10+s(NUM06,5)+CAT11+NUM07+CAT12+CAT13, + family = binomial(link = logit), data = train.data,na.action = na.exclude, + control = list(epsilon = 0.001,bf.epsilon = 0.001, maxit = 50, + bf.maxit = 10, trace = F)) Error in terms.formula(reformulate(term[i])) : invalid model formula in ExtractVars And after deleting df's model.gam - gam(formula = RES ~ CAT01+s(NUM01)+CAT02+CAT03+s(NUM02)+CAT04+ + CAT05+s(NUM03)+CAT06+CAT07+s(NUM04)+CAT08+s(NUM05)+CAT09+ + CAT10+s(NUM06)+CAT11+NUM07+CAT12+CAT13, + family = binomial(link = logit), data = train.data) Error in smooth.construct.tp.smooth.spec(object, data, knots) : A term has fewer unique covariate combinations than specified maximum degrees of freedom Can anybody show me some light in this case!!! Thanks in advance. -- View this message in context: http://www.nabble.com/Error-using-mgcv-package-tf3900783.html#a11058255 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] {spam?} Nonlinear Regression
Hi tronter, PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. Hank On Jun 7, 2007, at 5:50 PM, tronter wrote: Hello I followed the example in page 59, chapter 11 of the 'Introduction to R' manual. I entered my own x,y data. I used the least squares. My function has 5 parameters: p[1], p[2], p[3], p[4], p[5]. I plotted the x-y data. Then I used lines(spline(xfit,yfit)) to overlay best curves on the data while changing the parameters. My question is how do I calculate the residual sum of squares. In the example they have the following: df - data.frame( x=x, y=y) fit - nls(y ~SSmicmen(s, Vm, K), df) fit In the second line how would I input my function? Would it be: fit - nls(y ~ myfunction(p[1], p[2], p[3], p[4], p[5]), df) where myfunction is the actual function? My function doesnt have a name, so should I just enter it? Thanks -- View this message in context: http://www.nabble.com/Nonlinear- Regression-tf3886617.html#a11016968 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Rounding?
I was a bit puzzed by:
formatC(6.65,format=f,digits=1)
[1] 6.6
So I experimented and found:
formatC(6.6501,format=f,digits=1)
[1] 6.6
formatC(6.651,format=f,digits=1)
[1] 6.7
round(6.6501,1)
[1] 6.7
round(6.651,1)
[1] 6.7
version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status
major 2
minor 5.0
year 2007
month 04
day23
svn rev41293
language R
version.string R version 2.5.0 (2007-04-23)
My machine runs Windows NT.
Is this intended or just a Windows facility?
__
Bendix Carstensen
Senior Statistician
Steno Diabetes Center
Niels Steensens Vej 2-4
DK-2820 Gentofte
Denmark
+45 44 43 87 38 (direct)
+45 30 75 87 38 (mobile)
+45 44 43 73 13 (fax)
[EMAIL PROTECTED] http://www.biostat.ku.dk/~bxc
This e-mail (including any attachments) is intended for the ...{{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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Tools For Preparing Data For Analysis
Chris Evans wrote: Thanks Ted, great thread and I'm impressed with EpiData that I've discovered through this. I'd still like something that is even more integrated with R but maybe some day, if EpiData go fully open source as I think they are doing (A full conversion plan to secure this and convert the software to open-source has been made (See complete description of license and principles). at http://www.epidata.dk/ but the link to http://www.epidata.dk/about.htm doesn't exactly clarify this I don't think. But I can hope.) Thanks, yet again, to everyone who creates and contributes to the R system and this list: wonderful! Perhaps what we need is an XML standard for describing record-oriented data and its validation? This could then be used to validate a set of records and possibly also to build input forms with built-in validation for new records. You could then write R code that did 'check this data frame against this XML description and tell me the invalid rows'. Or Python code. This is the kind of thing that is traditionally built using a database front-end, but keeping the description in XML means that alternate interfaces (web forms, standalone programs using Qt or GTK libraries) can be used on the same description set. I had a quick search to see if this kind of thing exists already, but google searches for 'data entry verification' indicate that I should really pay some people in India to do that kind of thing for me... Barry __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Rounding?
your number 6.6501 is to large to fit in a floating point
number. It takes 56 bits and there are only 54 in a real number so the
system see it as 6.65 and does the rounding to an even digit; 6.6
6.651 does fit into a real number (takes 54 bits) and this will
now round to 6.7
On 6/11/07, BXC (Bendix Carstensen) [EMAIL PROTECTED] wrote:
I was a bit puzzed by:
formatC(6.65,format=f,digits=1)
[1] 6.6
So I experimented and found:
formatC(6.6501,format=f,digits=1)
[1] 6.6
formatC(6.651,format=f,digits=1)
[1] 6.7
round(6.6501,1)
[1] 6.7
round(6.651,1)
[1] 6.7
version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status
major 2
minor 5.0
year 2007
month 04
day23
svn rev41293
language R
version.string R version 2.5.0 (2007-04-23)
My machine runs Windows NT.
Is this intended or just a Windows facility?
__
Bendix Carstensen
Senior Statistician
Steno Diabetes Center
Niels Steensens Vej 2-4
DK-2820 Gentofte
Denmark
+45 44 43 87 38 (direct)
+45 30 75 87 38 (mobile)
+45 44 43 73 13 (fax)
[EMAIL PROTECTED] http://www.biostat.ku.dk/~bxc
This e-mail (including any attachments) is intended for the ...{{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
and provide commented, minimal, self-contained, reproducible code.
--
Jim Holtman
Cincinnati, OH
+1 513 646 9390
What is the problem you are trying to solve?
[[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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Rounding?
A good reference is:
*What Every Computer Scientist Should Know About Floating-Point*
*...*http://docs.sun.com/source/806-3568/ncg_goldberg.html
http://docs.sun.com/source/806-3568/ncg_goldberg.html
On 6/11/07, BXC (Bendix Carstensen) [EMAIL PROTECTED] wrote:
I was a bit puzzed by:
formatC(6.65,format=f,digits=1)
[1] 6.6
So I experimented and found:
formatC(6.6501,format=f,digits=1)
[1] 6.6
formatC(6.651,format=f,digits=1)
[1] 6.7
round(6.6501,1)
[1] 6.7
round(6.651,1)
[1] 6.7
version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status
major 2
minor 5.0
year 2007
month 04
day23
svn rev41293
language R
version.string R version 2.5.0 (2007-04-23)
My machine runs Windows NT.
Is this intended or just a Windows facility?
__
Bendix Carstensen
Senior Statistician
Steno Diabetes Center
Niels Steensens Vej 2-4
DK-2820 Gentofte
Denmark
+45 44 43 87 38 (direct)
+45 30 75 87 38 (mobile)
+45 44 43 73 13 (fax)
[EMAIL PROTECTED] http://www.biostat.ku.dk/~bxc
This e-mail (including any attachments) is intended for the ...{{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
and provide commented, minimal, self-contained, reproducible code.
--
Jim Holtman
Cincinnati, OH
+1 513 646 9390
What is the problem you are trying to solve?
[[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
and provide commented, minimal, self-contained, reproducible code.
[R] simultaneous computing
Hello, which possibilities are available in R for simultaneous or parallel computing? I only could find biopara (http://cran.r-project.org/src/contrib/Descriptions/biopara.html) Are there other possibilities? Are there special groups working on simultaneous computing with R? Thanks Markus -- Dipl.-Tech. Math. Markus Schmidberger Ludwig-Maximilians-Universität München IBE - Institut für medizinische Informationsverarbeitung, Biometrie und Epidemiologie __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Selecting all values smaller than X in a dataframe
Dear R users, I have a correlation matrix for a dataframe called synth, for which I now want to select only those cells that have correlations larger than +/-0.6: synth=data.frame(x=rnorm(10,1),y=rnorm(10,2),z=rnorm(10,0.5)) w=cor(synth,use=pairwise.complete.obs) w=as.data.frame(w) w[,sapply(w,abs(w),,0.6)] The problem is that using sapply with or doesn´t seem to work. How could I solve this problem? Thank you very much in advance for your help! Best wishes Christoph (I am using R 2.5.0 on Windows XP). -- Christoph Scherber DNPW, Agroecology University of Goettingen Waldweg 26 D-37073 Goettingen +49-(0)551-39-8807 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] simultaneous computing
There's RMPI tutorial: http://ace.acadiau.ca/math/ACMMaC/Rmpi/index.html On 11-Jun-07, at 9:11 AM, Markus Schmidberger wrote: Hello, which possibilities are available in R for simultaneous or parallel computing? I only could find biopara (http://cran.r-project.org/src/contrib/Descriptions/biopara.html) Are there other possibilities? Are there special groups working on simultaneous computing with R? Thanks Markus -- Dipl.-Tech. Math. Markus Schmidberger Ludwig-Maximilians-Universität München IBE - Institut für medizinische Informationsverarbeitung, Biometrie und Epidemiologie __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. -- Mike Lawrence Graduate Student, Department of Psychology, Dalhousie University Website: http://myweb.dal.ca/mc973993 Public calendar: http://icalx.com/public/informavore/Public The road to wisdom? Well, it's plain and simple to express: Err and err and err again, but less and less and less. - Piet Hein __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Recoding
I want to do some recoding of variables: code Age into groups and recode a factor into a smaller number of levels. There are a couple of options for recode functions, in the car package and in memisc, and I think in gmisc. Does anyone have any opinions on the the easiest, most reliable approach for these problems? David Scott _ David Scott Department of Statistics, Tamaki Campus The University of Auckland, PB 92019 Auckland 1142,NEW ZEALAND Phone: +64 9 373 7599 ext 86830 Fax: +64 9 373 7000 Email: [EMAIL PROTECTED] Graduate Officer, Department of Statistics Director of Consulting, Department of Statistics __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Selecting all values smaller than X in a dataframe
Christoph Scherber-2 wrote: Dear R users, I have a correlation matrix for a dataframe called synth, for which I now want to select only those cells that have correlations larger than +/-0.6: synth=data.frame(x=rnorm(10,1),y=rnorm(10,2),z=rnorm(10,0.5)) w=cor(synth,use=pairwise.complete.obs) w=as.data.frame(w) w[,sapply(w,abs(w),,0.6)] The problem is that using sapply with or doesn´t seem to work. How could I solve this problem? If you want to extract correlations with absolute value 0.6, then simply use w[abs(w)0.6] Please, reread the help(sapply). You give some extra arguments to this function. The first goes the vector, the second goes the function, and then - additional arguments to the function. Probably, you wanted w[sapply(abs(w),,0.6)] This gives the same result. -- View this message in context: http://www.nabble.com/Selecting-all-values-smaller-than-X-in-a-dataframe-tf3901238.html#a11059804 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code.
[R] Odp: Selecting all values smaller than X in a dataframe
Hi [EMAIL PROTECTED] napsal dne 11.06.2007 14:09:45: Dear R users, I have a correlation matrix for a dataframe called synth, for which I now want to select only those cells that have correlations larger than +/-0.6: synth=data.frame(x=rnorm(10,1),y=rnorm(10,2),z=rnorm(10,0.5)) w=cor(synth,use=pairwise.complete.obs) w=as.data.frame(w) Why? Better is tu use abs(w).6 and/or which(abs(w).6, arr.ind=T) or, if you want actual values just w[abs(w).6] Regards Petr w[,sapply(w,abs(w),,0.6)] The problem is that using sapply with or doesn´t seem to work. How could I solve this problem? Thank you very much in advance for your help! Best wishes Christoph (I am using R 2.5.0 on Windows XP). -- Christoph Scherber DNPW, Agroecology University of Goettingen Waldweg 26 D-37073 Goettingen +49-(0)551-39-8807 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] simultaneous computing
The package snow available from CRAN is one possibility. Best, luke On Mon, 11 Jun 2007, Markus Schmidberger wrote: Hello, which possibilities are available in R for simultaneous or parallel computing? I only could find biopara (http://cran.r-project.org/src/contrib/Descriptions/biopara.html) Are there other possibilities? Are there special groups working on simultaneous computing with R? Thanks Markus -- Luke Tierney Chair, Statistics and Actuarial Science Ralph E. Wareham Professor of Mathematical Sciences University of Iowa Phone: 319-335-3386 Department of Statistics andFax: 319-335-3017 Actuarial Science 241 Schaeffer Hall email: [EMAIL PROTECTED] Iowa City, IA 52242 WWW: http://www.stat.uiowa.edu __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Looking for R-code for non-negative matrix factorization in the presence of Gaussian or Poisson noise
On Mon, 11 Jun 2007, [EMAIL PROTECTED] wrote: Hi all, Has any of you implemented code for non-negative matrix factorization to solve Y=T P' +E; dim(Y)=n,p ; dim(T)=n,nc; dim (P)=(p,nc); dim(E)=n,p where T and P must be non-negative and E either Gaussian or Poisson noise. I'm looking for two variants: 1. Easy (I think), T is known (that is we just want to solve the general inverse problem) This is non-negative least squares, a quadratic programming problem, for which there is code (at least if n and nc are not too big) 2. Harder (?), T is unknown (under some restrictions) [as an intermediate, we may want to fix nc] Even with fixed nc this is Distinctly Non-trivial. It often isn't identifiable, for a start. I've encountered this problem in air pollution source apportionment, where people use an algorithm due to Paatero (1999) JCGS 8:854-8, which is a conjugate gradient algorithm that handles the constraints by creative abuse of preconditioning. The algorithm seems fairly well-behaved, although the statistical properties of the estimates are not well-understood [at least, I don't understand them, and I have simulations that appear to contradict the views of people who claim to understand them]. The difficulty probably depends on the size of the problem -- the air pollution problems have n~1000, p~20, nc~7, or larger. -thomas __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Selecting all values smaller than X in a dataframe
Christoph Scherber Christoph.Scherber at agr.uni-goettingen.de writes: Dear R users, I have a correlation matrix for a dataframe called synth, for which I now want to select only those cells that have correlations larger than +/-0.6: synth=data.frame(x=rnorm(10,1),y=rnorm(10,2),z=rnorm(10,0.5)) w=cor(synth,use=pairwise.complete.obs) w=as.data.frame(w) w[,sapply(w,abs(w),,0.6)] Dear Christoph, just change the last command to w[abs(w)0.6] Best wishes Ingo __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Recoding
I've used the one in car with no problems. For grouping a continuous variable you might also consider cut. On 11/06/07, David Scott [EMAIL PROTECTED] wrote: I want to do some recoding of variables: code Age into groups and recode a factor into a smaller number of levels. There are a couple of options for recode functions, in the car package and in memisc, and I think in gmisc. Does anyone have any opinions on the the easiest, most reliable approach for these problems? David Scott _ David Scott Department of Statistics, Tamaki Campus The University of Auckland, PB 92019 Auckland 1142,NEW ZEALAND Phone: +64 9 373 7599 ext 86830 Fax: +64 9 373 7000 Email: [EMAIL PROTECTED] Graduate Officer, Department of Statistics Director of Consulting, Department of Statistics __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- = David Barron Said Business School University of Oxford Park End Street Oxford OX1 1HP __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] biplot package II
Dear all,
I've been learning biplot (Gabriel, 1971) and some days ago I sent for
this list a procedural function with invitation for a collaborative package.
Jari Oksanen made some suggestions and I agree with all.
So, I reworked the function under the object-oriented programming
(OOP/S3). I think it is now a good frame for more resources.
Below it is the function and a small script to learn it:
#===
# Name : biplot.s
# Author : Jose Claudio Faria (DCET/USC/BRAZIL)
# Date (dd/mm/yy): 9/6/2007 13:33:48
# Version: v1.1
# Aim: 2d and 3d (under scaterplot3d and rgl packages) biplot
# Mail : [EMAIL PROTECTED]
#===
# Arguments:
# x Data (frame or matrix: objects in lines variables in columns)
# or a object of the class 'prcomp'.
# lambda.iniFirst eigenvalue to be considered (default is 1)
# lambda.endLatest eigenvalue to be considered
# (default is 2 to 2d or 3 to 3d)
# centerEither a logical value or a numeric vector of length equal
# to the number of columns of x (TRUE is the default).
# scale Either a logical value or a numeric vector of length equal
# to the number of columns of x (FALSE is the default).
# weightWay of factorize: 'equal', 'objects', 'variables'
# ('equal' is the default).
# plot Logical to produce or not a graphical representation of
# biplot (TRUE is the default).
# rgl.use If TRUE the 3d scatter will be under the rgl environment, in
# another way the scatterplot3d will be used ( the default).
# aspect3d Apparent ratios of the x, y, and z axes of the bounding box.
# clear3d Logical to clear or not a 3d graphical representation of
# biplot before to make a new (TRUE is the default).
# simple.axes Whether to draw simple axes (TRUE or FALSE).
# box Whether to draw a box (the default is FALSE).
# spheres Logical to represent objects as spheres (FALSE is the default).
# sphere.factor Relative size factor of sphere representing points; the
# default size is dependent on the scale of observations.
# col.obj Color of spheres or labels of objects.
# col.var Color of lines and labels of variables.
# var.factorFactor of expansion/reduction of length lines of the variables.
# graphical variables representation (=1, 1 is the default).
# cex Character expansion (for while valid only to graphics and
# scatterplot3d, not to rgl, packages).
#===
# Require 'rgl' and 'scatterplot3d' packages.
#===
# check the necessary packages
necessary = c('rgl', 'scatterplot3d')
if(!all(necessary %in% installed.packages()[, 'Package']))
install.packages(c('rgl', 'scatterplot3d'), dep = T)
# Plot 2d with 'graphics' packages
plot.biplot.2d = function(scores,
g,
hl,
lambda.ini,
lambda.end,
col.obj,
col.var,
var.factor,
cex)
{
plot(scores,
xlab=paste('PC', lambda.ini, sep=''),
ylab=paste('PC', lambda.end, sep=''),
type='n')
text(x=g[,1], y=g[,2],
labels=rownames(g),
cex=cex, col=col.obj)
arrows(x0=0, y0=0,
x1=hl[,1]*var.factor, y1=hl[,2]*var.factor,
length=0.1, angle=20,
col=col.var)
text(x=hl[,1]*var.factor, y=hl[,2]*var.factor,
labels = rownames(hl),
cex=cex, col=col.var)
}
# Plot 3d with 'scatterplot3d' package
plot.biplot.3d.default = function(scores,
g,
hl,
lambda.ini,
lambda.end,
col.obj,
col.var,
var.factor,
spheres,
box,
cex)
{
require(scatterplot3d)
graph = scatterplot3d(scores,
type = if(spheres) 'p' else 'n',
xlab=paste('PC', lambda.ini, sep=''),
ylab=paste('PC', lambda.ini+1, sep=''),
zlab=paste('PC', lambda.end, sep=''),
grid=F,
box=box,
cex.symbols=cex,
color=col.obj,
pch=20)
if(!spheres)
[R] Gini coefficient in R
If I use the Ineq library and the Gini function in this way: Gini(c(100,0,0,0)) I obtain the result 0.75 instead of 1 (that is the perfect inequality). I think Gini's formula in Ineq is based on a formula as reported here: http://mathworld.wolfram.com/GiniCoefficient.html but in the case of perfect inequality: x_1=...=x_n-1 =0 x_n0 these formula are equal to 1 - 1/n, not to 1. I don't know where I'm wrong -- Michele De Meo http://micheledemeo.blogspot.com/ [[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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] Gini coefficient in R
try this version instead:
gini - function(x, unbiased = TRUE, na.rm = FALSE){
if (!is.numeric(x)){
warning('x' is not numeric; returning NA)
return(NA)
}
if (!na.rm any(na.ind - is.na(x)))
stop('x' contain NAs)
if (na.rm)
x - x[!na.ind]
n - length(x)
mu - mean(x)
N - if (unbiased) n * (n - 1) else n * n
ox - x[order(x)]
dsum - drop(crossprod(2 * 1:n - n - 1, ox))
dsum / (mu * N)
}
gini(c(100,0,0,0))
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://med.kuleuven.be/biostat/
http://www.student.kuleuven.be/~m0390867/dimitris.htm
- Original Message -
From: MICHELE DE MEO [EMAIL PROTECTED]
To: r-help@stat.math.ethz.ch
Sent: Monday, June 11, 2007 4:13 PM
Subject: [R] Gini coefficient in R
If I use the Ineq library and the Gini function in this way:
Gini(c(100,0,0,0))
I obtain the result 0.75 instead of 1 (that is the perfect
inequality).
I think Gini's formula in Ineq is based on a formula as reported
here:
http://mathworld.wolfram.com/GiniCoefficient.html
but in the case of perfect inequality:
x_1=...=x_n-1 =0
x_n0
these formula are equal to 1 - 1/n, not to 1.
I don't know where I'm wrong
--
Michele De Meo
http://micheledemeo.blogspot.com/
[[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
and provide commented, minimal, self-contained, reproducible code.
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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Rounding?
jim holtman wrote: your number 6.6501 is to large to fit in a floating point number. It takes 56 bits and there are only 54 in a real number so the system see it as 6.65 and does the rounding to an even digit; 6.6 6.651 does fit into a real number (takes 54 bits) and this will now round to 6.7 Actually, a bit more insidious than that because 6.65 does not have an exact binary representation. Hence round(66.5) [1] 66 round(6.65,1) [1] 6.7 round(0.665,2) [1] 0.66 Notice that these are from Linux and differ from what you get on Windows. -- 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 and provide commented, minimal, self-contained, reproducible code.
[R] design package
Hi all: I tried to install design package for R used under the window. But it dose not work. Here is the message from R. library(Design) Error in library(Design) : 'Design' is not a valid package -- installed 2.0.0? I did not find 2.x version of Design package for window 95. I installed one of two zip files available in the internet. Please let me know what's wrong? Thanks, Qin [[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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] p-value from GEE why factor 2*pnorm?
I got an answer for the other question (thank you) But there is another question (I am afraid this is a basic question ...) In this tread there is a hint hwo to calculate the p-vlue of an GEE: _http://finzi.psych.upenn.edu/R/Rhelp02a/archive/74150.html_ Then, get the P values using a normal approximation for the distribution of z: / 2 * pnorm(abs(coef(summary(fm1))[,5]), lower.tail = FALSE) / (Intercept) TPTLD 0. 0.04190831 1. why is the result multiplicated with 2? There is a P-value between 1 and 2 with the results below and multiplicated with 2: 2*pnorm(c(1.8691945,0.5882351,2.4903091,1.9287802,2.3172983,2.2092593,2.2625959,1.6395695), lower.tail =TRUE) Regards Carmen __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] p-value from GEE why factor 2*pnorm?
the recommendation was to use lower.tail=FALSE. b On Jun 11, 2007, at 11:21 AM, Carmen Meier wrote: I got an answer for the other question (thank you) But there is another question (I am afraid this is a basic question ...) In this tread there is a hint hwo to calculate the p-vlue of an GEE: _http://finzi.psych.upenn.edu/R/Rhelp02a/archive/74150.html_ Then, get the P values using a normal approximation for the distribution of z: / 2 * pnorm(abs(coef(summary(fm1))[,5]), lower.tail = FALSE) / (Intercept) TPTLD 0. 0.04190831 1. why is the result multiplicated with 2? There is a P-value between 1 and 2 with the results below and multiplicated with 2: 2*pnorm(c (1.8691945,0.5882351,2.4903091,1.9287802,2.3172983,2.2092593,2.2625959 ,1.6395695), lower.tail =TRUE) Regards Carmen __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] p-value from GEE why factor 2*pnorm?
Benilton Carvalho schrieb: the recommendation was to use lower.tail=FALSE. b O but then the results are significant and this does not match the observation. The results are matching the observations if the formula is pnorm(c(1.8691945,0.5882351,2.4903091,1.9287802,2.3172983,2.2092593,2.2625959,1.6395695), lower.tail =TRUE) so I have any unknown problem anywhere :-( REgards Carmen __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] lmekin() function in kinship package
Hi, I had a problem with the lmekin() in kinship package: lmekin() can not be wrapped into another function library(kinship) #creat an example dataset xx-rnorm(100) yy-rnorm(100) id-1:100 test.dat-as.data.frame(cbind(xx,yy,id)) rm(xx,yy,id) a-bdsmatrix(rep(10,10),rep(block,10),dimnames=list(c(1:100),c(1:100))) #100x100 block (n=10) diagonal matrix to indicate the correlation between the 100 observations #this works, call lmekin directly lmekin(yy~xx,random=~1|id,data=test.dat,varlist=a,na.action=na.omit) #this doesn't work, wrap into another function fo-function(x,y,z,data)lmekin(y~x,random=~1|z,data=test.dat,varlist=a,na.action=na.omit) fo(xx,yy,id) Error in eval(expr, envir, enclos) : Object y not found I did line by line debug within lmekin(). I found the problem was from line #25 m - eval(m, sys.parent()) Here is the print(m) result before this line [1] model.frame(data = data, na.action = na.omit, formula = y ~ x +z) I will appreciate anyone help me figure out why data cannot be passed to the arguments when lmekin() is wrapped into another function? And what can be done to fix it? Thanks Qiong __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] autoregressive spectral density estimate by andrews' plug-in method?
Hello! I would like to ask if there is in R a function that estimates the spectral density function of a stochastic series at frequency zero by the plug-in method, advocated by Andrews in his paper Heteroscedasticity and Autocorrelation Consistent Covariance Matrix Estimation, Econometrica, 59,817-858. I saw R has functions that employ Andrews' plug-in method using an AR(1) approximation for the estimation of the variance-covariance matrix in linear models. They come with the sandwich package. The so called meat is actually the estimate of the spectral density matrix of the model coefficients at frequency zero. However, I have a time series of length 160 and I need to estimate its spectral density via Andrews methodology. Any suggestions will be appreciated. Excuse me if I am asking something obvious. Regards, Martin - Нека музиката бъде с теб! http://musicidol.btv.bg/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] open .r files with double-click
However, do note (on Windows) that you can use an external text/programming editors (see CRAN's listings)and can register .r / .R files to open automatically in the chosen editor when clicked on.At least some of these editors (eg TINN-R) can be configured to automatically and simultaneously open the RGUI, too, I believe -- but someone may correct me on this. Bert Gunter Nonclinical Statistics 7-7374 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Duncan Murdoch Sent: Saturday, June 09, 2007 4:29 AM To: [EMAIL PROTECTED] Cc: r-help@stat.math.ethz.ch Subject: Re: [R] open .r files with double-click On 08/06/2007 2:52 PM, [EMAIL PROTECTED] wrote: Hi Folks, On Windows XP, R 2.5.0. After reading the Installation for Windows and Windows FAQs, I cannot resolve this. I set file types so that Rgui.exe will open .r files. When I try to open a .r file by double-clicking, R begins to launch, but I get an error message saying Argument 'C:\Documents and Settings\Zoology\My Documents\trial.r' _ignored_ I click OK, and then R GUI opens, but not the script file. Is there a way to change this? Not currently. See the appendix Invoking R of the Introduction manual for the current command line parameters, which don't include open a script. This would be a reasonable addition, and I'll add it at some point, sooner if someone else comes up with a convincing argument for the right command line parameter to do this. It would be better if clicking on a second script opened a new window in the same session, but that takes more work; not sure I'll get to this. 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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] p-value from GEE why factor 2*pnorm?
At 11:21 AM 6/11/2007, Carmen wrote: snip In this tread there is a hint hwo to calculate the p-vlue of an GEE: _http://finzi.psych.upenn.edu/R/Rhelp02a/archive/74150.html Then, get the P values using a normal approximation for the distribution of z: / 2 * pnorm(abs(coef(summary(fm1))[,5]), lower.tail = FALSE) / (Intercept) TPTLD 0. 0.04190831 1. why is the result multiplicated with 2? There is a P-value between 1 and 2 with the results below and multiplicated with 2: 2*pnorm(c(1.8691945,0.5882351,2.4903091,1.9287802,2.3172983,2.2092593,2.2625959,1.6395695), lower.tail =TRUE) 1. The given in the thread mentioned was: 2 * pnorm(abs(coef(summary(fm1))[,5]), lower.tail = FALSE) 2. The reason for the 2 at the front is to make it an equal-tails or 2-sided confidence interval. Pedantically, you should use 1.96 instead of 2.0 for consistency, but 2.0 = 1.96 rounded to one decimal place. 3. This is what is usually called a Wald type confidence interval, as it is simply the normal quantile (+/- 1.96) multiplied by the standard error of estimate to get the +/- widths for the interval. These would be added to the estimate itself to get the final Wald confidence interval, which obviously assumes a normal distribution applies. Robert A. LaBudde, PhD, PAS, Dpl. ACAFS e-mail: [EMAIL PROTECTED] Least Cost Formulations, Ltd.URL: http://lcfltd.com/ 824 Timberlake Drive Tel: 757-467-0954 Virginia Beach, VA 23464-3239Fax: 757-467-2947 Vere scire est per causas scire __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Spencer,
Thank you for your response. I also did not see anything on the lsoda help page
which is the reason that I wrote to the list.
From your response, I am not sure if I asked my question clearly.
I am modeling a group of people (in a variety of health states) moving through
time (and getting infected with an infectious disease). This means that the
count of the number of people in each state should be positive at all times.
What appears to happen is that lsoda asks for a derivative at a given point in
time t and then adjusts the state of the population. However, perhaps due to
numerical instability, it occasionally lower the population count below 0 for
one of the health states (perhaps because it's step size is too big or
something).
I have tried both the logarithm trick and also changing the relative and
absolute tolerance inputs but I still get the problem for certain combinations
of parameters and initial conditions.
It occurs both under MS Windows XP Service Pack 2 and on a Linux cluster so I
am pretty sure it is not platform specific.
My real question to the group is if there is not a work around in lsoda are
there other ode solvers in R that will allow the constraint of solutions to the
ODEs remain non-negative?
Best regards,
Jeremy
Spencer Graves [EMAIL PROTECTED] 6/8/2007 9:51 AM
On the 'lsoda' help page, I did not see any option to force some
or all parameters to be nonnegative.
Have you considered replacing the parameters that must be
nonnegative with their logarithms? This effective moves the 0 lower
limit to (-Inf) and seems to have worked well for me in the past.
Often, it can even make the log likelihood or sum of squares surface
more elliptical, which means that the standard normal approximation for
the sampling distribution of parameter estimates will likely be more
accurate.
Hope this helps.
Spencer Graves
p.s. Your example seems not to be self contained. If I could have
easily copied it from your email and run it myself, I might have been
able to offer more useful suggestions.
Jeremy Goldhaber-Fiebert wrote:
Hello,
I am using odesolve to simulate a group of people moving through time and
transmitting infections to one another.
In Matlab, there is a NonNegative option which tells the Matlab solver to
keep the vector elements of the ODE solution non-negative at all times. What
is the right way to do this in R?
Thanks,
Jeremy
P.S., Below is a simplified version of the code I use to try to do this, but
I am not sure that it is theoretically right
dynmodel - function(t,y,p)
{
## Initialize parameter values
birth - p$mybirth(t)
death - p$mydeath(t)
recover - p$myrecover
beta - p$mybeta
vaxeff - p$myvaxeff
vaccinated - p$myvax(t)
vax - vaxeff*vaccinated/100
## If the state currently has negative quantities (shouldn't have), then
reset to reasonable values for computing meaningful derivatives
for (i in 1:length(y)) {
if (y[i]0) {
y[i] - 0
}
}
S - y[1]
I - y[2]
R - y[3]
N - y[4]
shat - (birth*(1-vax)) - (death*S) - (beta*S*I/N)
ihat - (beta*S*I/N) - (death*I) - (recover*I)
rhat - (birth*(vax)) + (recover*I) - (death*R)
## Do we overshoot into negative space, if so shrink derivative to bring
state to 0
## then rescale the components that take the derivative negative
if (shat+S0) {
shat_old - shat
shat - -1*S
scaled_transmission - (shat/shat_old)*(beta*S*I/N)
ihat - scaled_transmission - (death*I) - (recover*I)
}
if (ihat+I0) {
ihat_old - ihat
ihat - -1*I
scaled_recovery - (ihat/ihat_old)*(recover*I)
rhat - scaled_recovery +(birth*(vax)) - (death*R)
}
if (rhat+R0) {
rhat - -1*R
}
nhat - shat + ihat + rhat
if (nhat+N0) {
nhat - -1*N
}
## return derivatives
list(c(shat,ihat,rhat,nhat),c(0))
}
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Re: [R] design package
Which version of R For R-2.5.0, Design (with D, not d) does not pass the checks under Windows. Hence there is no Windows version for R-2.5.0 currently available on CRAN. For older versions of R (down to R-1.7.x): Please either use the Windows binary version, r.g. by typing install.packages(Design) of the package or compile from sources yourself. Look like you just uncompressed some source package without installing it. Qin Qin wrote: Hi all: I tried to install design package for R used under the window. But it dose not work. Here is the message from R. library(Design) Error in library(Design) : 'Design' is not a valid package -- installed 2.0.0? The error message means R, not the package. I did not find 2.x version of Design package for window 95. I installed one of two zip files available in the internet. Is the really Windows 95? Uwe Ligges Please let me know what's wrong? Thanks, Qin [[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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] p-value from GEE why factor 2*pnorm?
Well, AFAIK, the definition of a p-value is the probability of observing something at least as extreme as the observed data. If you observed z, and Z follows a std-normal p-value = P( Z -abs(z) ) + P( Z abs(z) ) = 2*P ( Z abs(z) ) = 2*pnorm(z, lower.tail=FALSE) try z=0 (you should get 1) and z=1.96 (you should get 5%) b On Jun 11, 2007, at 11:34 AM, Carmen Meier wrote: Benilton Carvalho schrieb: the recommendation was to use lower.tail=FALSE. b O but then the results are significant and this does not match the observation. The results are matching the observations if the formula is pnorm(c (1.8691945,0.5882351,2.4903091,1.9287802,2.3172983,2.2092593,2.2625959 ,1.6395695), lower.tail =TRUE) so I have any unknown problem anywhere :-( REgards Carmen __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] how to ignore error messages?
Dear group: I wrote a code to iterate a non-linear fit with a set of data. The entire process didn't implemented to the end because an error message, singular gradient. I knew that some sub-sets (columns) do not fit my formula well and may result in parameters going to infinity. It is pretty hard to remove those sub sets before running the code since that will take a lot of time. I added some logic assessments prior to running nonlinear fit. It helped but some exceptions still existed. I am wondering if there is any way in R by which I can continue to run the entire code to the end by ignoring the error message? Greatly appreciate your help, Jianping ## Jianping Jin Ph.D. Bioinformatics scientist Center for Bioinformatics Room 3133 Bioinformatics building CB# 7104 University of Chapel Hill Chapel Hill, NC 27599 Phone: (919)843-6105 FAX: (919)843-3103 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 and provide commented, minimal, self-contained, reproducible code.
[R] history in R.app on Mac
hi, I am really confused by history function in R.app for Mac: Here is some test from command-line: ls() character(0) ls() character(0) ls() character(0) history() ls() ls() ls() history() That is what I expected. But from R.app, it does not show anything. _ platform i386-apple-darwin8.9.1 arch i386 os darwin8.9.1 system i386, darwin8.9.1 status major 2 minor 5.0 year 2007 month 04 day23 svn rev41293 language R version.string R version 2.5.0 (2007-04-23) Thanks, -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. Did you always know? No, I did not. But I believed... ---Matrix III __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Jeremy,
First, setting hmax to a small number could prevent a large step, if
you think that is a problem. Second, however, I don't see how you can
get a negative population size when using the log trick. I would
think that that would prevent completely any negative values of N
(i.e. e^-10 0). Can you explain? or do you want to a void that
trick? The only other solver I know of is rk4 and it is not recommended.
Hank
On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote:
Hi Spencer,
Thank you for your response. I also did not see anything on the
lsoda help page which is the reason that I wrote to the list.
From your response, I am not sure if I asked my question clearly.
I am modeling a group of people (in a variety of health states)
moving through time (and getting infected with an infectious
disease). This means that the count of the number of people in each
state should be positive at all times.
What appears to happen is that lsoda asks for a derivative at a
given point in time t and then adjusts the state of the population.
However, perhaps due to numerical instability, it occasionally
lower the population count below 0 for one of the health states
(perhaps because it's step size is too big or something).
I have tried both the logarithm trick and also changing the
relative and absolute tolerance inputs but I still get the problem
for certain combinations of parameters and initial conditions.
It occurs both under MS Windows XP Service Pack 2 and on a Linux
cluster so I am pretty sure it is not platform specific.
My real question to the group is if there is not a work around in
lsoda are there other ode solvers in R that will allow the
constraint of solutions to the ODEs remain non-negative?
Best regards,
Jeremy
Spencer Graves [EMAIL PROTECTED] 6/8/2007 9:51 AM
On the 'lsoda' help page, I did not see any option to force some
or all parameters to be nonnegative.
Have you considered replacing the parameters that must be
nonnegative with their logarithms? This effective moves the 0 lower
limit to (-Inf) and seems to have worked well for me in the past.
Often, it can even make the log likelihood or sum of squares surface
more elliptical, which means that the standard normal approximation
for
the sampling distribution of parameter estimates will likely be more
accurate.
Hope this helps.
Spencer Graves
p.s. Your example seems not to be self contained. If I could have
easily copied it from your email and run it myself, I might have been
able to offer more useful suggestions.
Jeremy Goldhaber-Fiebert wrote:
Hello,
I am using odesolve to simulate a group of people moving through
time and transmitting infections to one another.
In Matlab, there is a NonNegative option which tells the Matlab
solver to keep the vector elements of the ODE solution non-
negative at all times. What is the right way to do this in R?
Thanks,
Jeremy
P.S., Below is a simplified version of the code I use to try to do
this, but I am not sure that it is theoretically right
dynmodel - function(t,y,p)
{
## Initialize parameter values
birth - p$mybirth(t)
death - p$mydeath(t)
recover - p$myrecover
beta - p$mybeta
vaxeff - p$myvaxeff
vaccinated - p$myvax(t)
vax - vaxeff*vaccinated/100
## If the state currently has negative quantities (shouldn't
have), then reset to reasonable values for computing meaningful
derivatives
for (i in 1:length(y)) {
if (y[i]0) {
y[i] - 0
}
}
S - y[1]
I - y[2]
R - y[3]
N - y[4]
shat - (birth*(1-vax)) - (death*S) - (beta*S*I/N)
ihat - (beta*S*I/N) - (death*I) - (recover*I)
rhat - (birth*(vax)) + (recover*I) - (death*R)
## Do we overshoot into negative space, if so shrink derivative to
bring state to 0
## then rescale the components that take the derivative negative
if (shat+S0) {
shat_old - shat
shat - -1*S
scaled_transmission - (shat/shat_old)*(beta*S*I/N)
ihat - scaled_transmission - (death*I) - (recover*I)
}
if (ihat+I0) {
ihat_old - ihat
ihat - -1*I
scaled_recovery - (ihat/ihat_old)*(recover*I)
rhat - scaled_recovery +(birth*(vax)) - (death*R)
}
if (rhat+R0) {
rhat - -1*R
}
nhat - shat + ihat + rhat
if (nhat+N0) {
nhat - -1*N
}
## return derivatives
list(c(shat,ihat,rhat,nhat),c(0))
}
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-
guide.html
and provide commented, minimal, self-contained,
Re: [R] how to ignore error messages?
see ?try Jianping Jin wrote: Dear group: I wrote a code to iterate a non-linear fit with a set of data. The entire process didn't implemented to the end because an error message, singular gradient. I knew that some sub-sets (columns) do not fit my formula well and may result in parameters going to infinity. It is pretty hard to remove those sub sets before running the code since that will take a lot of time. I added some logic assessments prior to running nonlinear fit. It helped but some exceptions still existed. I am wondering if there is any way in R by which I can continue to run the entire code to the end by ignoring the error message? Greatly appreciate your help, Jianping ## Jianping Jin Ph.D. Bioinformatics scientist Center for Bioinformatics Room 3133 Bioinformatics building CB# 7104 University of Chapel Hill Chapel Hill, NC 27599 Phone: (919)843-6105 FAX: (919)843-3103 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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] how to ignore error messages?
Jianping Jin jjin at email.unc.edu writes: ... I wrote a code to iterate a non-linear fit with a set of data. The entire process didn't implemented to the end because an error message, singular gradient. I knew that some sub-sets (columns) do not fit my formula well and may result in parameters going to infinity. It is pretty hard to remove those sub sets before running the code since that will take a lot of time. I added some logic assessments prior to running nonlinear fit. It helped but some exceptions still existed. I am wondering if there is any way in R by which I can continue to run the entire code to the end by ignoring the error message? ... ?try Or, even easier, use nlsList in nlme, which does exactly what you want, handling convergence errors graciously. 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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] history in R.app on Mac
If you click on the history icon in the toolbar in R.app you will see them. Or up and down arrows. R.app implements some extra features, such as multiline command retrieval. Through preference settings you can control if you want to see just a single ls() or multiple, etc. Rob On Jun 11, 2007, at 9:18 AM, Weiwei Shi wrote: hi, I am really confused by history function in R.app for Mac: Here is some test from command-line: ls() character(0) ls() character(0) ls() character(0) history() ls() ls() ls() history() That is what I expected. But from R.app, it does not show anything. _ platform i386-apple-darwin8.9.1 arch i386 os darwin8.9.1 system i386, darwin8.9.1 status major 2 minor 5.0 year 2007 month 04 day23 svn rev41293 language R version.string R version 2.5.0 (2007-04-23) Thanks, -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. Did you always know? No, I did not. But I believed... ---Matrix III __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
in line Martin Henry H. Stevens wrote: Hi Jeremy, First, setting hmax to a small number could prevent a large step, if you think that is a problem. Second, however, I don't see how you can get a negative population size when using the log trick. SG: Can lsoda estimate complex or imaginary parameters? I would think that that would prevent completely any negative values of N (i.e. e^-10 0). Can you explain? or do you want to a void that trick? The only other solver I know of is rk4 and it is not recommended. Hank On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote: Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick snip __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Spencer, I have copied Woody Setzer. I have no idea whether lsoda can estimate parameters that could take imaginary values. Hank On Jun 11, 2007, at 12:52 PM, Spencer Graves wrote: in line Martin Henry H. Stevens wrote: Hi Jeremy, First, setting hmax to a small number could prevent a large step, if you think that is a problem. Second, however, I don't see how you can get a negative population size when using the log trick. SG: Can lsoda estimate complex or imaginary parameters? Hmm. I have no idea. I would think that that would prevent completely any negative values of N (i.e. e^-10 0). Can you explain? or do you want to a void that trick? The only other solver I know of is rk4 and it is not recommended. Hank On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote: Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick snip __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. Dr. Hank Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/~stevenmh/ http://www.muohio.edu/ecology/ http://www.muohio.edu/botany/ E Pluribus Unum __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] history in R.app on Mac
got it. Thanks. I was always trying to find it at menu and did not expect that in my toolbar :) I just assumed that any tool icon should also be shown on menu, which is just a routine to me. -w On 6/11/07, Rob J Goedman [EMAIL PROTECTED] wrote: If you click on the history icon in the toolbar in R.app you will see them. Or up and down arrows. R.app implements some extra features, such as multiline command retrieval. Through preference settings you can control if you want to see just a single ls() or multiple, etc. Rob On Jun 11, 2007, at 9:18 AM, Weiwei Shi wrote: hi, I am really confused by history function in R.app for Mac: Here is some test from command-line: ls() character(0) ls() character(0) ls() character(0) history() ls() ls() ls() history() That is what I expected. But from R.app, it does not show anything. _ platform i386-apple-darwin8.9.1 arch i386 os darwin8.9.1 system i386, darwin8.9.1 status major 2 minor 5.0 year 2007 month 04 day23 svn rev41293 language R version.string R version 2.5.0 (2007-04-23) Thanks, -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. Did you always know? No, I did not. But I believed... ---Matrix III __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. Did you always know? No, I did not. But I believed... ---Matrix III __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Problem with RSVGTipsDevice
Hi there.
I am still trying to get the RSVGTipsDevice to work, yet I can not.
I have copied the first example from RSVGTipsDevice documentation:
library(RSVGTipsDevice)
devSVGTips(C:\\svgplot1.svg, toolTipMode=1,
title=SVG example plot 1: shapes and points, tooltips are title + 1 line)
plot(c(0,10),c(0,10), type=n, xlab=x, ylab=y,
main=Example SVG plot with title + 1 line tips (mode=1))
setSVGShapeToolTip(title=A rectangle, desc=that is yellow)
rect(1,1,4,6, col='yellow')
setSVGShapeToolTip(title=1st circle with title only)
points(5.5,7.5,cex=20,pch=19,col='red')
setSVGShapeToolTip(title=A triangle, desc=big and green)
polygon(c(3,6,8), c(3,6,3), col='green')
# no tooltips on these points
points(2:8, 8:2, cex=3, pch=19, col='black')
# tooltips on each these points
invisible(sapply(1:7, function(x)
{setSVGShapeToolTip(title=paste(point, x))
points(x+1, 8-x, cex=3, pch=1, col='black')}))
dev.off()
This results in the following output:
http://www.nabble.com/file/p11064573/svgplot1.svg svgplot1.svg
It opens but when I try and hover over the triangle, for example, I do not
get a topptip box appear. I have tried opening the file though firefox, and
XP IE - and on more than one computer yet it does not work. Do I need to
install something else as well?
Many thanks
--
View this message in context:
http://www.nabble.com/Problem-with-RSVGTipsDevice-tf3902760.html#a11064573
Sent from the R help mailing list archive at Nabble.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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Weighted least squares
As John noted, there are different kinds of weights, and different terminology: * inverse-variance weights (accuracy weights) * case weights (frequencies, counts) * sampling weights (selection probability weights) I'll add: * inverse-variance weights, where var(y for observation) = 1/weight (as opposed to just being inversely proportional to the weight) * weights used as part of an algorithm (e.g. for robust estimation, or glm's using iteratively-reweighted least-squares). For linear regression, the type of weights don't affect regression coefficient calculation, but do affect inferences such as standard errors for the regression coefficients, degrees of freedom for variance estimates, etc. lm() inferences assume the first type. Other formulae are appropriate for inferences for types 2-4. Combinations of types 1-4 require other formulae; this gets nontrivial. For the 5th type, inferences need to be handled by the algorithm that is using weighted linear regression. Tim Hesterberg John Fox wrote: I think that the problem is that the term weights has different meanings, which, although they are related, are not quite the same. The weights used by lm() are (inverse-)variance weights, reflecting the variances of the errors, with observations that have low-variance errors therefore being accorded greater weight in the resulting WLS regression. What you have are sometimes called case weights, and I'm unaware of a general way of handling them in R, although you could regenerate the unaggregated data. As you discovered, you get the same coefficients with case weights as with variance weights, but different standard errors. Finally, there are sampling weights, which are inversely proportional to the probability of selection; these are accommodated by the survey package. To complicate matters, this terminology isn't entirely standard. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Spencer, Lsoda does not estimate any parameters (nlmeODE does parameter estimation). It just computes the solution trajectory, at discrete times, of a dynamical systems (i.e. set of differential equations). It only works with real numbers, as far as I know. Ravi. --- Ravi Varadhan, Ph.D. Assistant Professor, The Center on Aging and Health Division of Geriatric Medicine and Gerontology Johns Hopkins University Ph: (410) 502-2619 Fax: (410) 614-9625 Email: [EMAIL PROTECTED] Webpage: http://www.jhsph.edu/agingandhealth/People/Faculty/Varadhan.html -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Spencer Graves Sent: Monday, June 11, 2007 12:53 PM To: Martin Henry H. Stevens Cc: Jeremy Goldhaber-Fiebert; r-help@stat.math.ethz.ch Subject: Re: [R] Fwd: Using odesolve to produce non-negative solutions in line Martin Henry H. Stevens wrote: Hi Jeremy, First, setting hmax to a small number could prevent a large step, if you think that is a problem. Second, however, I don't see how you can get a negative population size when using the log trick. SG: Can lsoda estimate complex or imaginary parameters? I would think that that would prevent completely any negative values of N (i.e. e^-10 0). Can you explain? or do you want to a void that trick? The only other solver I know of is rk4 and it is not recommended. Hank On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote: Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick snip __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Problem with RSVGTipsDevice
Hi Mister_Bluesman,
Sorry to sadden you further, but your example svg works perfectly on my
machine (R 2.5 running on Vista). All the ToolTips display. Perhaps it's
your browser [setup]: I'm using Opera 9.20 (Build 8771).
HTH,
Mark Difford.
mister_bluesman wrote:
Hi there.
I am still trying to get the RSVGTipsDevice to work, yet I can not.
I have copied the first example from RSVGTipsDevice documentation:
library(RSVGTipsDevice)
devSVGTips(C:\\svgplot1.svg, toolTipMode=1,
title=SVG example plot 1: shapes and points, tooltips are title + 1
line)
plot(c(0,10),c(0,10), type=n, xlab=x, ylab=y,
main=Example SVG plot with title + 1 line tips (mode=1))
setSVGShapeToolTip(title=A rectangle, desc=that is yellow)
rect(1,1,4,6, col='yellow')
setSVGShapeToolTip(title=1st circle with title only)
points(5.5,7.5,cex=20,pch=19,col='red')
setSVGShapeToolTip(title=A triangle, desc=big and green)
polygon(c(3,6,8), c(3,6,3), col='green')
# no tooltips on these points
points(2:8, 8:2, cex=3, pch=19, col='black')
# tooltips on each these points
invisible(sapply(1:7, function(x)
{setSVGShapeToolTip(title=paste(point, x))
points(x+1, 8-x, cex=3, pch=1, col='black')}))
dev.off()
This results in the following output:
http://www.nabble.com/file/p11064573/svgplot1.svg svgplot1.svg
It opens but when I try and hover over the triangle, for example, I do not
get a topptip box appear. I have tried opening the file though firefox,
and XP IE - and on more than one computer yet it does not work. Do I need
to install something else as well?
Many thanks
--
View this message in context:
http://www.nabble.com/Problem-with-RSVGTipsDevice-tf3902760.html#a11065021
Sent from the R help mailing list archive at Nabble.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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Problem with RSVGTipsDevice
Mister_Bluesman,
Perhaps I should have been more precise: your included svgplot1.svg displays
fine...!
mister_bluesman wrote:
Hi there.
I am still trying to get the RSVGTipsDevice to work, yet I can not.
I have copied the first example from RSVGTipsDevice documentation:
library(RSVGTipsDevice)
devSVGTips(C:\\svgplot1.svg, toolTipMode=1,
title=SVG example plot 1: shapes and points, tooltips are title + 1
line)
plot(c(0,10),c(0,10), type=n, xlab=x, ylab=y,
main=Example SVG plot with title + 1 line tips (mode=1))
setSVGShapeToolTip(title=A rectangle, desc=that is yellow)
rect(1,1,4,6, col='yellow')
setSVGShapeToolTip(title=1st circle with title only)
points(5.5,7.5,cex=20,pch=19,col='red')
setSVGShapeToolTip(title=A triangle, desc=big and green)
polygon(c(3,6,8), c(3,6,3), col='green')
# no tooltips on these points
points(2:8, 8:2, cex=3, pch=19, col='black')
# tooltips on each these points
invisible(sapply(1:7, function(x)
{setSVGShapeToolTip(title=paste(point, x))
points(x+1, 8-x, cex=3, pch=1, col='black')}))
dev.off()
This results in the following output:
http://www.nabble.com/file/p11064573/svgplot1.svg svgplot1.svg
It opens but when I try and hover over the triangle, for example, I do not
get a topptip box appear. I have tried opening the file though firefox,
and XP IE - and on more than one computer yet it does not work. Do I need
to install something else as well?
Many thanks
--
View this message in context:
http://www.nabble.com/Problem-with-RSVGTipsDevice-tf3902760.html#a11065061
Sent from the R help mailing list archive at Nabble.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
and provide commented, minimal, self-contained, reproducible code.
[R] debugger library. Error?
Hi, running the following example code (taken from the docs) will prodice an error in my system (R 2.5.0, win): library(debug) mtrace(glm) # turns tracing on names( tracees) # glm check.for.tracees( package:base) # glm glm(stupid.args) # voila le debugger Error in all.levs[[j]] : subscript out of bounds It happens with any function, I just posted one form the example. Thanks, -J -- Jose Quesada, PhD. http://www.andrew.cmu.edu/~jquesada __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Jeremy,
A smaller step size may or may not help. If the issue is simply truncation
error, that is the error involved in discretizing the differential
equations, then a smaller step size would help. If, however, the true
solution to the differential equation is negative, for some t, then the
numerical solution should also be negative. If the negative solution does
not make sense, then the system of equation needs to be examined to see when
and why negative solutions arise. Perhaps, I am just making this up - there
needs to be a dampening function that slows down the trajectory as it
approaches zero from its initial value. It is also possible that only
certain regions of the parameter space (note that initial conditions are
also parameters) are allowed in the sense that only there the solution is
feasible for all t. So, in your example, the parameters might not be
realistic. In short, if you are sure that the numerical solution is
accurate, then you need to go back to your system of equations and analyze
them carefully.
Ravi.
---
Ravi Varadhan, Ph.D.
Assistant Professor, The Center on Aging and Health
Division of Geriatric Medicine and Gerontology
Johns Hopkins University
Ph: (410) 502-2619
Fax: (410) 614-9625
Email: [EMAIL PROTECTED]
Webpage: http://www.jhsph.edu/agingandhealth/People/Faculty/Varadhan.html
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jeremy
Goldhaber-Fiebert
Sent: Monday, June 11, 2007 11:47 AM
To: Spencer Graves
Cc: r-help@stat.math.ethz.ch
Subject: Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Spencer,
Thank you for your response. I also did not see anything on the lsoda help
page which is the reason that I wrote to the list.
From your response, I am not sure if I asked my question clearly.
I am modeling a group of people (in a variety of health states) moving
through time (and getting infected with an infectious disease). This means
that the count of the number of people in each state should be positive at
all times.
What appears to happen is that lsoda asks for a derivative at a given point
in time t and then adjusts the state of the population. However, perhaps due
to numerical instability, it occasionally lower the population count below 0
for one of the health states (perhaps because it's step size is too big or
something).
I have tried both the logarithm trick and also changing the relative and
absolute tolerance inputs but I still get the problem for certain
combinations of parameters and initial conditions.
It occurs both under MS Windows XP Service Pack 2 and on a Linux cluster so
I am pretty sure it is not platform specific.
My real question to the group is if there is not a work around in lsoda are
there other ode solvers in R that will allow the constraint of solutions to
the ODEs remain non-negative?
Best regards,
Jeremy
Spencer Graves [EMAIL PROTECTED] 6/8/2007 9:51 AM
On the 'lsoda' help page, I did not see any option to force some
or all parameters to be nonnegative.
Have you considered replacing the parameters that must be
nonnegative with their logarithms? This effective moves the 0 lower
limit to (-Inf) and seems to have worked well for me in the past.
Often, it can even make the log likelihood or sum of squares surface
more elliptical, which means that the standard normal approximation for
the sampling distribution of parameter estimates will likely be more
accurate.
Hope this helps.
Spencer Graves
p.s. Your example seems not to be self contained. If I could have
easily copied it from your email and run it myself, I might have been
able to offer more useful suggestions.
Jeremy Goldhaber-Fiebert wrote:
Hello,
I am using odesolve to simulate a group of people moving through time and
transmitting infections to one another.
In Matlab, there is a NonNegative option which tells the Matlab solver to
keep the vector elements of the ODE solution non-negative at all times. What
is the right way to do this in R?
Thanks,
Jeremy
P.S., Below is a simplified version of the code I use to try to do this,
but I am not sure that it is theoretically right
dynmodel - function(t,y,p)
{
## Initialize parameter values
birth - p$mybirth(t)
death - p$mydeath(t)
recover - p$myrecover
beta - p$mybeta
vaxeff - p$myvaxeff
vaccinated - p$myvax(t)
vax - vaxeff*vaccinated/100
## If the state currently has negative quantities (shouldn't have), then
reset to reasonable values for computing meaningful derivatives
for (i in 1:length(y)) {
if (y[i]0) {
y[i] - 0
}
}
S - y[1]
I - y[2]
R - y[3]
N - y[4]
shat -
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Ravi, Thanks for your response. I tried this in Berkeley Madonna and in Matlab. In Berkeley Madonna I did not have the problem (RK4 solver). In Matlab (ode45 solver), I had the problem if I did not use their NonNegative option. My thought was that NonNegative uses something like an additional piece of logic in modifying step size (maybe something like: if stepsize * derivative + current condition is negative, then reduce step size), but I don't know. My original application was to generate ode output for a variety of unknown parameters, comparing the output to observed data and thereby using a likelihood-based approach to identify unknown parameter combinations that are most consistent with observed data given the uncertainty. First, I wanted to get a sense of how the model performed over a range of parameters that seem plausible based on literature review. I had originally thought to do this in Matlab and built a little proof of concept searching program, but given that R is free and has many other great packages for doing optimization and statistical analysis (and is easy to setup which is great for cluster computing situations), I thought it would be better to do it in R. When I did this, I got negative values and hence sent to the list. Once again, thanks for your help. Not sure if this clarification email will generate other suggestions or thoughts. Best, Jeremy Ravi Varadhan [EMAIL PROTECTED] 6/11/2007 2:11 PM Hi Jeremy, A smaller step size may or may not help. If the issue is simply truncation error, that is the error involved in discretizing the differential equations, then a smaller step size would help. If, however, the true solution to the differential equation is negative, for some t, then the numerical solution should also be negative. If the negative solution does not make sense, then the system of equation needs to be examined to see when and why negative solutions arise. Perhaps, I am just making this up - there needs to be a dampening function that slows down the trajectory as it approaches zero from its initial value. It is also possible that only certain regions of the parameter space (note that initial conditions are also parameters) are allowed in the sense that only there the solution is feasible for all t. So, in your example, the parameters might not be realistic. In short, if you are sure that the numerical solution is accurate, then you need to go back to your system of equations and analyze them carefully. Ravi. --- Ravi Varadhan, Ph.D. Assistant Professor, The Center on Aging and Health Division of Geriatric Medicine and Gerontology Johns Hopkins University Ph: (410) 502-2619 Fax: (410) 614-9625 Email: [EMAIL PROTECTED] Webpage: http://www.jhsph.edu/agingandhealth/People/Faculty/Varadhan.html -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jeremy Goldhaber-Fiebert Sent: Monday, June 11, 2007 11:47 AM To: Spencer Graves Cc: r-help@stat.math.ethz.ch Subject: Re: [R] Fwd: Using odesolve to produce non-negative solutions Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick and also changing the relative and absolute tolerance inputs but I still get the problem for certain combinations of parameters and initial conditions. It occurs both under MS Windows XP Service Pack 2 and on a Linux cluster so I am pretty sure it is not platform specific. My real question to the group is if there is not a work around in lsoda are there other ode solvers in R that will allow the constraint of solutions to the ODEs remain non-negative? Best regards, Jeremy Spencer Graves [EMAIL PROTECTED] 6/8/2007 9:51 AM On the 'lsoda' help page, I did not see any option to force some or all parameters to be nonnegative. Have you considered replacing the parameters that must be nonnegative with their logarithms? This effective moves the 0 lower limit to (-Inf) and seems to have worked well for me in the past. Often, it can even make the log likelihood or sum of squares surface more
Re: [R] Problem with RSVGTipsDevice
Ah. Now that's intersting. It works in Opera. But do you get an annoying
'Null' label by the cursor when you place it over the svg file?
MANY thanks
Mark Difford wrote:
Mister_Bluesman,
Perhaps I should have been more precise: your included svgplot1.svg
displays fine...!
mister_bluesman wrote:
Hi there.
I am still trying to get the RSVGTipsDevice to work, yet I can not.
I have copied the first example from RSVGTipsDevice documentation:
library(RSVGTipsDevice)
devSVGTips(C:\\svgplot1.svg, toolTipMode=1,
title=SVG example plot 1: shapes and points, tooltips are title + 1
line)
plot(c(0,10),c(0,10), type=n, xlab=x, ylab=y,
main=Example SVG plot with title + 1 line tips (mode=1))
setSVGShapeToolTip(title=A rectangle, desc=that is yellow)
rect(1,1,4,6, col='yellow')
setSVGShapeToolTip(title=1st circle with title only)
points(5.5,7.5,cex=20,pch=19,col='red')
setSVGShapeToolTip(title=A triangle, desc=big and green)
polygon(c(3,6,8), c(3,6,3), col='green')
# no tooltips on these points
points(2:8, 8:2, cex=3, pch=19, col='black')
# tooltips on each these points
invisible(sapply(1:7, function(x)
{setSVGShapeToolTip(title=paste(point, x))
points(x+1, 8-x, cex=3, pch=1, col='black')}))
dev.off()
This results in the following output:
http://www.nabble.com/file/p11064573/svgplot1.svg svgplot1.svg
It opens but when I try and hover over the triangle, for example, I do
not get a topptip box appear. I have tried opening the file though
firefox, and XP IE - and on more than one computer yet it does not work.
Do I need to install something else as well?
Many thanks
--
View this message in context:
http://www.nabble.com/Problem-with-RSVGTipsDevice-tf3902760.html#a11065835
Sent from the R help mailing list archive at Nabble.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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
By the way, if someone could forward the original question to me (I'm subscribed to but not currently receiving R-help, as I found I was spending too much time reading it!) I might think of something more useful. (alternatively, when was it posted; I can find it on gmane, too). Woody R. Woodrow Setzer, Ph. D. National Center for Computational Toxicology US Environmental Protection Agency Mail Drop B205-01/US EPA/RTP, NC 27711 Ph: (919) 541-0128Fax: (919) 541-1194 Martin Henry H. Stevens [EMAIL PROTECTED]To .eduSpencer Graves [EMAIL PROTECTED] 06/11/2007 01:02cc PM Jeremy Goldhaber-Fiebert [EMAIL PROTECTED], R-Help r-help@stat.math.ethz.ch, Woodrow Setzer/RTP/USEPA/[EMAIL PROTECTED] Subject Re: [R] Fwd: Using odesolve to produce non-negative solutions Hi Spencer, I have copied Woody Setzer. I have no idea whether lsoda can estimate parameters that could take imaginary values. Hank On Jun 11, 2007, at 12:52 PM, Spencer Graves wrote: in line Martin Henry H. Stevens wrote: Hi Jeremy, First, setting hmax to a small number could prevent a large step, if you think that is a problem. Second, however, I don't see how you can get a negative population size when using the log trick. SG: Can lsoda estimate complex or imaginary parameters? Hmm. I have no idea. I would think that that would prevent completely any negative values of N (i.e. e^-10 0). Can you explain? or do you want to a void that trick? The only other solver I know of is rk4 and it is not recommended. Hank On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote: Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick snip __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. Dr. Hank Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/~stevenmh/ http://www.muohio.edu/ecology/ http://www.muohio.edu/botany/ E Pluribus Unum __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi, all. lsoda can certainly not handle complex parameters. You can try (as Hank suggested) limiting hmax. You can also crank up relative and absolute precision by specifying smaller values of rtol and atol. I've seen similar problems in which the state variable becomes negative, with very small absolute value, when theoretically, the system has a non-negative solution. This is certainly due to imprecision in the numerical solution. Have you tried including an analytic jacobian? That could improve the numeric properties of the solution. Woody R. Woodrow Setzer, Ph. D. National Center for Computational Toxicology US Environmental Protection Agency Mail Drop B205-01/US EPA/RTP, NC 27711 Ph: (919) 541-0128Fax: (919) 541-1194 Martin Henry H. Stevens [EMAIL PROTECTED]To .eduSpencer Graves [EMAIL PROTECTED] 06/11/2007 01:02cc PM Jeremy Goldhaber-Fiebert [EMAIL PROTECTED], R-Help r-help@stat.math.ethz.ch, Woodrow Setzer/RTP/USEPA/[EMAIL PROTECTED] Subject Re: [R] Fwd: Using odesolve to produce non-negative solutions Hi Spencer, I have copied Woody Setzer. I have no idea whether lsoda can estimate parameters that could take imaginary values. Hank On Jun 11, 2007, at 12:52 PM, Spencer Graves wrote: in line Martin Henry H. Stevens wrote: Hi Jeremy, First, setting hmax to a small number could prevent a large step, if you think that is a problem. Second, however, I don't see how you can get a negative population size when using the log trick. SG: Can lsoda estimate complex or imaginary parameters? Hmm. I have no idea. I would think that that would prevent completely any negative values of N (i.e. e^-10 0). Can you explain? or do you want to a void that trick? The only other solver I know of is rk4 and it is not recommended. Hank On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote: Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick snip __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. Dr. Hank Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/~stevenmh/ http://www.muohio.edu/ecology/ http://www.muohio.edu/botany/ E Pluribus Unum __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Problem with RSVGTipsDevice
Hi Mister_Bluesman, Sadly it is; and it stays up (in what's probably a bold, sans serif), lurking behind the ToolTip, when that comes up. This is not my area of X, but any standard txt editor will open an svg file. If you search through your file you will find lots and lots of null arguments to various functions. It's likely that the problem lies there. HTH, Mark Difford. mister_bluesman wrote: Ah. Now that's intersting. It works in Opera. But do you get an annoying 'Null' label by the cursor when you place it over the svg file? MANY thanks -- View this message in context: http://www.nabble.com/Problem-with-RSVGTipsDevice-tf3902760.html#a11066065 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code.
[R] R vs. Splus in Pharma/Devices Industry
Following up to some extent on Friday's discussion regarding the 'validation' of R, could I ask the list group's opinion on possible advantages of R over Splus from a pharma/devices perspective? I wish to exclude the obvious price difference, which doesnât seem to carry as much weight as I would have thought. Besides, I have noticed many former Splus users gravitating towards R, and I suspect that the reasons are not purely economic. I can think of a few advantages of Splus: 1. SeqTrial (of course that means more $) 2. Tech support 3. The warm fuzzies that management seems to get from proprietary software I can also think of a few advantages of R: 1. Based on my personal experiences, simulations requiring a lot of looping seem to run faster. 2. R interfaces with BUGS, for example through BRUGS. 3. The wonderful help list! As always, I am speaking for myself and not necessarily for Edwards Lifesciences. Regards, -Cody Cody Hamilton, PhD Edwards Lifesciences [[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 and provide commented, minimal, self-contained, reproducible code.
[R] Reading old S-plus dmp files
I'm sorry that this question has been asked before but I ask it again because in the archives I didn't see a solution. It's an old S-plus dmp file for a hierarchical bayes linear model program written by DuMouchel and available publicly and freely at: ftp://ftp.research.att.com/dist/bayes-meta/hblm.dmp Only problem is that it cannot be read in R, even using the functions of the library foreign to read dmp files. Does anyone know a workaround, maybe hints about what parts of the file need to be modified so that it can be read. I have read the documentations of the program, it seems a very interesting program, it's a shame that I can't read it. Thanks. -- View this message in context: http://www.nabble.com/Reading-old-S-plus-dmp-files-tf3903236.html#a11066203 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
[EMAIL PROTECTED] wrote: Following up to some extent on Friday's discussion regarding the 'validation' of R, could I ask the list group's opinion on possible advantages of R over Splus from a pharma/devices perspective? I wish to exclude the obvious price difference, which doesn’t seem to carry as much weight as I would have thought. Besides, I have noticed many former Splus users gravitating towards R, and I suspect that the reasons are not purely economic. I can think of a few advantages of Splus: 1. SeqTrial (of course that means more $) 2. Tech support 3. The warm fuzzies that management seems to get from proprietary software I can also think of a few advantages of R: 1. Based on my personal experiences, simulations requiring a lot of looping seem to run faster. 2. R interfaces with BUGS, for example through BRUGS. 3. The wonderful help list! As always, I am speaking for myself and not necessarily for Edwards Lifesciences. Regards, -Cody Cody Hamilton, PhD Edwards Lifesciences [[alternative HTML version deleted]] A big one for us is plotmath (for clinical trial reports we put a lot of greek letters and subscripts on plots), and later we will consider migrating a lot of our stuff to the ggplot package which I don't think is available in S-Plus. Lexical scoping is another advantage of R as is the ability to reference files on the internet. Frank __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt 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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Jeremy, A smaller step size may or may not help. If the issue is simply truncation error, that is the error involved in discretizing the differential equations, then a smaller step size would help. If, however, the true solution to the differential equation is negative, for some t, then the numerical solution should also be negative. If the negative solution does not make sense, then the system of equation needs to be examined to see when and why negative solutions arise. Perhaps, I am just making this up - there needs to be a barrier function that slows down the trajectory as it approaches zero from its initial value. It is also possible that only certain regions of the parameter space are allowed in the sense that only there the solution is feasible for all t. So, in your example, the parameters might not be realistic. In short, if you are sure that the numerical solution is accurate, then you need to go back to your system of equations and analyze them carefully. Ravi. --- Ravi Varadhan, Ph.D. Assistant Professor, The Center on Aging and Health Division of Geriatric Medicine and Gerontology Johns Hopkins University Ph: (410) 502-2619 Fax: (410) 614-9625 Email: [EMAIL PROTECTED] Webpage: http://www.jhsph.edu/agingandhealth/People/Faculty/Varadhan.html -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Martin Henry H. Stevens Sent: Monday, June 11, 2007 1:03 PM To: Spencer Graves Cc: Jeremy Goldhaber-Fiebert; R-Help; [EMAIL PROTECTED] Subject: Re: [R] Fwd: Using odesolve to produce non-negative solutions Hi Spencer, I have copied Woody Setzer. I have no idea whether lsoda can estimate parameters that could take imaginary values. Hank On Jun 11, 2007, at 12:52 PM, Spencer Graves wrote: in line Martin Henry H. Stevens wrote: Hi Jeremy, First, setting hmax to a small number could prevent a large step, if you think that is a problem. Second, however, I don't see how you can get a negative population size when using the log trick. SG: Can lsoda estimate complex or imaginary parameters? Hmm. I have no idea. I would think that that would prevent completely any negative values of N (i.e. e^-10 0). Can you explain? or do you want to a void that trick? The only other solver I know of is rk4 and it is not recommended. Hank On Jun 11, 2007, at 11:46 AM, Jeremy Goldhaber-Fiebert wrote: Hi Spencer, Thank you for your response. I also did not see anything on the lsoda help page which is the reason that I wrote to the list. From your response, I am not sure if I asked my question clearly. I am modeling a group of people (in a variety of health states) moving through time (and getting infected with an infectious disease). This means that the count of the number of people in each state should be positive at all times. What appears to happen is that lsoda asks for a derivative at a given point in time t and then adjusts the state of the population. However, perhaps due to numerical instability, it occasionally lower the population count below 0 for one of the health states (perhaps because it's step size is too big or something). I have tried both the logarithm trick snip __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. Dr. Hank Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/~stevenmh/ http://www.muohio.edu/ecology/ http://www.muohio.edu/botany/ E Pluribus Unum __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Problem with RSVGTipsDevice
Ah. Now that's intersting. It works in Opera. But do you get an annoying 'Null' label by the cursor when you place it over the svg file? MANY thanks -- View this message in context: http://www.nabble.com/Problem-with-RSVGTipsDevice-tf3902760.html#a11065042 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code.
[R] Can I access the filename of the active editor window?
When I run a script from an open editor window (using Ctrl-A, Ctrl-R), I would like the filename of the script to be automatically written into the program output, to keep up with frequent version changes. Is there a way to access the filename (+ path) of the open script (the active one, if there is more than one editor window open)? I'm using R 2.4.1 on Windows XP. Paul [[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 and provide commented, minimal, self-contained, reproducible code.
[R] Rearranging Capture History Data in R
What code can i use to convert a table like this: Tag#Date 1 1 2 1 3 1 4 1 2 2 4 2 1 3 2 3 4 4 Into one like this: Tag 1 2 3 4 #Date header 1 1 0 0 1 2 1 1 1 0 3 1 0 0 0 4 1 1 0 1 Thanks, Ben Cox Research Assistant (M.S.) Montana Cooperative Fishery Research Unit 301 Lewis Hall Montana State University Bozeman, MT 59717 (406)994-6643 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] A Question about R
Hi Sir/Madam, I'm a researcher in university of Guelph, Canada and now considering using R to do some data analysis. I'm wondering whether there is a library available in R that includes algorithms for archetypal analysis? This is a method quite similar to principal components analysis that is designed to find archetypes or pure types from multidimensional data. Any help from you is greatly appreciated. Thanks Juan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Lines in dotchart dotplot ?
On 6/11/07, John Kane [EMAIL PROTECTED] wrote: --- [EMAIL PROTECTED] wrote: On 6/9/07, John Kane [EMAIL PROTECTED] wrote: Is it possible to use dotchart or dotplot and set the lines in such a way that they only extend from the left y-axis to the data point? Yes (sort of) in dotplot at least. E.g., dotplot(VADeaths, groups = FALSE, type = c(p, h)) dotplot(VADeaths, groups = FALSE, type = c(p, h), origin = 0) -Deepayan Ah, that is quite nice, not exactly what I remember from Cleveland but it should do quite nicely. You can do that too if you really want to: dotplot(VADeaths, groups = FALSE, type = c(p, h), origin = 0, panel = panel.xyplot, pch = 16, lty = 3, col = black) -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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fwd: Using odesolve to produce non-negative solutions
Jeremy,
You should examine the steady-state solution to your system of equations, by
setting the time-derivatives to zero and then solving/analyzing the
resulting algebraic equations. This should give you some insights.
Let us say you have 3 groups, A,B, and C, with initial conditions:
N_A(t=0) = N_{A0}, N_B(t=0) = N_{B0}, and N_C(t=0) = N_{C0}, and that people
transition in and out of these 3 states (one of the states could even be
absorbing, e.g. death), but it is true for any time t that N_A(t) + N_B(t) +
N_C(t) = N_{A0} + N_{B0} + N_{C0}. Furthermore, you have 3 diff-equations
that describe the rate of change of N_A, N_B, and N_C, for t 0. If it
happens that one of the N's, say N_A, becomes negative, you could set it
equal to zero. But then you have to figure out how to re-adjust N_B and N_C
so that they add up to the initial total count. After re-adjustment, you
also have to think about whether the resulting system of equations are
valid, when there are no A people.
Ravi.
---
Ravi Varadhan, Ph.D.
Assistant Professor, The Center on Aging and Health
Division of Geriatric Medicine and Gerontology
Johns Hopkins University
Ph: (410) 502-2619
Fax: (410) 614-9625
Email: [EMAIL PROTECTED]
Webpage: http://www.jhsph.edu/agingandhealth/People/Faculty/Varadhan.html
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jeremy
Goldhaber-Fiebert
Sent: Monday, June 11, 2007 11:47 AM
To: Spencer Graves
Cc: r-help@stat.math.ethz.ch
Subject: Re: [R] Fwd: Using odesolve to produce non-negative solutions
Hi Spencer,
Thank you for your response. I also did not see anything on the lsoda help
page which is the reason that I wrote to the list.
From your response, I am not sure if I asked my question clearly.
I am modeling a group of people (in a variety of health states) moving
through time (and getting infected with an infectious disease). This means
that the count of the number of people in each state should be positive at
all times.
What appears to happen is that lsoda asks for a derivative at a given point
in time t and then adjusts the state of the population. However, perhaps due
to numerical instability, it occasionally lower the population count below 0
for one of the health states (perhaps because it's step size is too big or
something).
I have tried both the logarithm trick and also changing the relative and
absolute tolerance inputs but I still get the problem for certain
combinations of parameters and initial conditions.
It occurs both under MS Windows XP Service Pack 2 and on a Linux cluster so
I am pretty sure it is not platform specific.
My real question to the group is if there is not a work around in lsoda are
there other ode solvers in R that will allow the constraint of solutions to
the ODEs remain non-negative?
Best regards,
Jeremy
Spencer Graves [EMAIL PROTECTED] 6/8/2007 9:51 AM
On the 'lsoda' help page, I did not see any option to force some
or all parameters to be nonnegative.
Have you considered replacing the parameters that must be
nonnegative with their logarithms? This effective moves the 0 lower
limit to (-Inf) and seems to have worked well for me in the past.
Often, it can even make the log likelihood or sum of squares surface
more elliptical, which means that the standard normal approximation for
the sampling distribution of parameter estimates will likely be more
accurate.
Hope this helps.
Spencer Graves
p.s. Your example seems not to be self contained. If I could have
easily copied it from your email and run it myself, I might have been
able to offer more useful suggestions.
Jeremy Goldhaber-Fiebert wrote:
Hello,
I am using odesolve to simulate a group of people moving through time and
transmitting infections to one another.
In Matlab, there is a NonNegative option which tells the Matlab solver to
keep the vector elements of the ODE solution non-negative at all times. What
is the right way to do this in R?
Thanks,
Jeremy
P.S., Below is a simplified version of the code I use to try to do this,
but I am not sure that it is theoretically right
dynmodel - function(t,y,p)
{
## Initialize parameter values
birth - p$mybirth(t)
death - p$mydeath(t)
recover - p$myrecover
beta - p$mybeta
vaxeff - p$myvaxeff
vaccinated - p$myvax(t)
vax - vaxeff*vaccinated/100
## If the state currently has negative quantities (shouldn't have), then
reset to reasonable values for computing meaningful derivatives
for (i in 1:length(y)) {
if (y[i]0) {
y[i] - 0
}
}
S - y[1]
I - y[2]
R - y[3]
N - y[4]
shat - (birth*(1-vax)) - (death*S) - (beta*S*I/N)
ihat -
Re: [R] Rearranging Capture History Data in R
date = c(1, 1, 1, 1, 2, 2, 3, 3, 4) tag = c(1, 2, 3, 4, 2, 4, 1, 2, 4) table(factor(tag, levels=1:4), factor(date, levels=1:4)) (not sure how you got Tag 1/Date 4 = 1) On Jun 11, 2007, at 3:11 PM, [EMAIL PROTECTED] wrote: What code can i use to convert a table like this: Tag#Date 1 1 2 1 3 1 4 1 2 2 4 2 1 3 2 3 4 4 Into one like this: Tag 1 2 3 4 #Date header 1 1 0 0 1 2 1 1 1 0 3 1 0 0 0 4 1 1 0 1 Thanks, Ben Cox Research Assistant (M.S.) Montana Cooperative Fishery Research Unit 301 Lewis Hall Montana State University Bozeman, MT 59717 (406)994-6643 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting- guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] simultaneous computing
Dear Markus, You might want to check Rmpi, papply, snow, rpvm, and nws. Best, R. On 6/11/07, Markus Schmidberger [EMAIL PROTECTED] wrote: Hello, which possibilities are available in R for simultaneous or parallel computing? I only could find biopara (http://cran.r-project.org/src/contrib/Descriptions/biopara.html) Are there other possibilities? Are there special groups working on simultaneous computing with R? Thanks Markus -- Dipl.-Tech. Math. Markus Schmidberger Ludwig-Maximilians-Universität München IBE - Institut für medizinische Informationsverarbeitung, Biometrie und Epidemiologie __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Ramon Diaz-Uriarte Statistical Computing Team Structural Biology and Biocomputing Programme Spanish National Cancer Centre (CNIO) http://ligarto.org/rdiaz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] selecting characters from a line of text
Is there a way in R to select certain characters from a line of text? I have some data that is presently in a large number of text files, and I would like to be able to select elements of each text file (elements are always on the same line, in the same position) and organize them into a table. Is there a tool to select text in this way in R? What I am looking for would be somewhat similar to the left() and right() functions in Excel. I have looked at the parse() and scan() functions, but don't think they can do what I want (although I could be wrong). Thank you, Tim [[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 and provide commented, minimal, self-contained, reproducible code.
[R] ylim settings
Hi, I need to know what is the plus/minus adjustment (proportion) that ylim applies if I have something like this: ylim=range(c(x,z)) meaning what's the x-(proportion) and the z+ (proportion)? Thank you, Judith Luggage? GPS? Comic books? __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Package update announcements
Hi Folks, I was wondering what everyone thought about adding a sentence to each package update announcement that described what the package did. R extensions are so numerous that it is difficult to keep up with them. Would it be appropriate to ask package developers to add a brief sentence about what the package does, when they announce updates? I would benefit from such descriptions. Cheers, Hank Dr. Hank Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/~stevenmh/ http://www.muohio.edu/ecology/ http://www.muohio.edu/botany/ E Pluribus Unum __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Overlaying lattice graphs
Hello
I apologize in advance if this question has already be posted on the
list, although I could not find a relevant thread in the archives.
I would like to overlay xyplots using different datasets for each plot.
I typically work on the following data.frame (mydata) structure
mydata
DrugTimeObserved Predicted
1 A0.05 10 10.2
2 A0.10 20 19.5
etc...
100 B0.05 11 12.7
101 B0.10 35 36
etc...
I want to plot the observed data as points and the predicted values as
lines. If I use the following commands, I don't have the possibility to
switch the y values from Observed for the scatterplot to Predicted for
the line.
xyplot(Observed ~ Time | Drug, data = mydata, panel = function(x,y, ...){
+panel.xyplot(x,y,...)
+panel.xyplot(x,y,type=l,...)})
I wonder if this problem can be solved using the trellis.focus family
commands but I have a hard time to understand how they work.
Please, let me know if a thread have already addressed this question.
Otherwise, I would grateful for any hint, comments or info you can provide.
Thanks
Sebastien
__
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Re: [R] ylim settings
On Mon, 2007-06-11 at 13:57 -0700, Judith Flores wrote: Hi, I need to know what is the plus/minus adjustment (proportion) that ylim applies if I have something like this: ylim=range(c(x,z)) meaning what's the x-(proportion) and the z+ (proportion)? Thank you, Judith If I am correctly understanding your query, the answer is +/- 4% of the range of values for the concatenated vectors, as you have it expressed above, _IF_ par(yaxs) is set to the default value of 'r'. If you set par(yaxs = i), then the y axis range is set to the min and max values of the concatenated vectors. See ?par for more information, specifically 'xaxs'. HTH, Marc Schwartz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Overlaying lattice graphs
On 6/11/07, Seb [EMAIL PROTECTED] wrote:
Hello
I apologize in advance if this question has already be posted on the
list, although I could not find a relevant thread in the archives.
I would like to overlay xyplots using different datasets for each plot.
I typically work on the following data.frame (mydata) structure
mydata
DrugTimeObserved Predicted
1 A0.05 10 10.2
2 A0.10 20 19.5
etc...
100 B0.05 11 12.7
101 B0.10 35 36
etc...
I want to plot the observed data as points and the predicted values as
lines. If I use the following commands, I don't have the possibility to
switch the y values from Observed for the scatterplot to Predicted for
the line.
xyplot(Observed ~ Time | Drug, data = mydata, panel = function(x,y, ...){
+panel.xyplot(x,y,...)
+panel.xyplot(x,y,type=l,...)})
I wonder if this problem can be solved using the trellis.focus family
commands but I have a hard time to understand how they work.
Please, let me know if a thread have already addressed this question.
Otherwise, I would grateful for any hint, comments or info you can provide.
There are several possible solutions. In your case, the simplest one
would be something like (see ?panel.superpose for explanation):
xyplot(Observed + Predicted ~ Time | Drug, data = mydata,
type = c(p, l), distribute.type = TRUE)
This will work best if the Time values are ordered; otherwise you could use
type = c(p, a)
instead, which will be a little slower. Let us know if this doesn't
give you what you want, preferably with a reproducible example
illustrating why.
-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
and provide commented, minimal, self-contained, reproducible code.
[R] Rgobbi and colours question
Hi I am using R to implement a multidimensional algorithm which maps places based on the distances between each other. This is presented really well in Rggobi. However, for each place plotted, I have another numerical statistic between 0 and 1 which I would like to represent by colouring the points in Rgobbi. However I do not know how to do this. The file I am using is called places.txt and has the following contents: Chelt Exeter London Birm Chelt 0 118 96 50 Exeter 118 0 118 163 London 96 118 0 118 Birm50 163 118 0 To plot these, the R code is as follows: library(rggobi) Places-read.table(C:\\places.txt) Places.location - cmdscale(Places, k=2) round(Places.location,0) g-ggobi(Places.location) Now, for each point plotted I want to assign a colour to it based on the statistic in a file called PlacesStats.txt which currently has the following format: 0.1 0.3 0.4 0.5 With the values in order of the names of the places given above so that: 0.1 is assigned to Chelt 0.3 is assigned to Exeter 0.4 is assigned to London 0.7 is assigned to Birm Ideally, I want a gradient of two colours so that say, 0 = blue and 1 = red and values between are a mixture. Any ideas on how I could do this? Would I need to change the PlacesStats.txt file so that is reads: Chelt 0.1 Exeter 0.3 London 0.4 Birm 0.7 for example? I would be very grateful if anybody could help. Many thanks -- View this message in context: http://www.nabble.com/Rgobbi-and-colours-question-tf3904622.html#a11070461 Sent from the R help mailing list archive at Nabble.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 and provide commented, minimal, self-contained, reproducible code.
Re: [R] selecting characters from a line of text
Maybe substring() is what you're looking for? Some examples: substring(textstring,1,5) [1] texts substring(textstring,3) [1] xtstring substring(textstring,3,nchar(textstring)) [1] xtstring --- Tim Holland [EMAIL PROTECTED] wrote: Is there a way in R to select certain characters from a line of text? I have some data that is presently in a large number of text files, and I would like to be able to select elements of each text file (elements are always on the same line, in the same position) and organize them into a table. Is there a tool to select text in this way in R? What I am looking for would be somewhat similar to the left() and right() functions in Excel. I have looked at the parse() and scan() functions, but don't think they can do what I want (although I could be wrong). Thank you, Tim [[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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] barplot and map overlay
Hi, I wonder if it is possible with the graphics package to overlay one or several plots (barplots, for example) over a map. Data for the map is in a data frame with the latitude and longitude coordinates, and then: plot(map$long, map$lat, type =l) produces the map. I want to put each barplot in specific locations on the map, namely at the center of statistical squares. I´ve seen an example in Paul Murrell´s R Graphics book that seems appropriate (grid package), but a bit complicated. Thanks a lot for any advice. Héctor -- Héctor Villalobos [EMAIL PROTECTED] CICIMAR - IPN A.P. 592. Col. Centro La Paz, Baja California Sur, MÉXICO. 23000 Tels. (+52 612) 122 53 44; 123 46 58; 123 47 34 ext. 2425 Fax. (+52 612) 122 53 22 [[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 and provide commented, minimal, self-contained, reproducible code.
[R] Components in the control list of optim()
Because the function optim() does not return the values of the components in the 'control' list, I am seeking help to uncover the blackbox for some of these components: parscale: par/parscale is used for ndeps, but how does optim() set parscale at the first place? abstol: no default information given in the help file, so what is it? reltol: this one is clear in the help file, but if both abstol and reltol are used as default, which one does optim() take as the higher stopping criterion? Two other related questions: First, ndeps is used for the finite-difference approximation, but is this forward, backward, or central difference? Second, if I minimize a function through optim (say, BFGS), does abstol/reltol use the function value or the estimated parameter values to assess convergence? Many thanks, Daniel Pinpoint customers who are looking for what you sell. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Unable to load RMySQL
All, I have been down for the past two weeks moving from Tulsa to Houston for a new job. I just got internet access yesterday at our new home. I will be working on the RMySQL binary later this week if possible. There is a hitch in compiling the binary. I have the binaries for an older version of MySQL on which the R compilation worked. A newer version of MySQL 5.0.37, I never was able to get the R binaries to compile. I appreciate your patience here because I have a new job and my family and I are trying to organize our new home right now. Thank you Joe Henric Nilsson (Public) wrote: Den Ti, 2007-02-13, 11:43 skrev Ravi S. Shankar: Hi R users, I am unable to load RMySQL. The zip file is not available which I guess is needed to load this pakage. Please read http://cran.r-project.org/bin/windows/contrib/2.4/ReadMe to find out why. I also tried extracting the package from RMySQL_0.5-11.tar.gz and then pasted the package in the directory where R is loaded for which I am getting the following error message Error in library(RMySQL) : 'RMySQL' is not a valid package -- installed 2.0.0? Did you really expect that to work? Any help would be welcome Joe Byers (http://bus.cba.utulsa.edu/byersj/Research.asp) kindly provides a Windows binary of RMySQL. HTH, Henric Thank you, Ravi [[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 and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Textpad help
Thx for the response John and Jim.
I had heard before that textpad is a good editor for R codes and one can run
the R code from textpad itself and once we do that, the output is displayed
on the textpad itself. The reason I wanted to use textpad is to make use of
the syntax highlighting a la SAS.
So I installed textpad and according to the instructions given in
http://www.data-for-all.com/blog/?p=68 . everything worked out fine until
I ran the following test code:
*setwd('C:/Temp/SampleR')
#
N-250
x-1:N
y-0.53*x +rnorm(x, 0.42*N, 0.14*N)
model1-lm(y~x)
print(summary(model1)) *
Once I submitted the R code using the SUBMIT R FILE in TOOLS from the
toolbar menu the following message displayed.
The filename, directory name, or volume label syntax is incorrect.
Tool completed with exit code 1
Now even If I run a simple code like *a-10;a* still the same message is
displayed.
I am kind of sure it is nothing wrong with the code but something is wrong
with the way the textpad installation. Any help? I am not a techie person.
Thx for your help
On 6/11/07, Mary Royerr [EMAIL PROTECTED] wrote:
I have installed textpad and tried running R code. But it gives me the
following error message.
The filename, directory name, or volume label syntax is incorrect.
Tool completed with exit code 1
Can you provide any help? I am not a technical person. So the help in
detail will be appreciated.
Thx
[[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
and provide commented, minimal, self-contained, reproducible code.
[R] question about data availale in .RData file using the biobase package
Hi all I am analyzing micro array data and I have R workspace images as my source of the data(.Rdata format).That was in the biobase package format,so I used some commands from the bio base package manual and could write the data into excel files. The data I am working on is the cancer data. I could get microarray information and recurrence information by using commands like x-pData(oncogene) y-exprs(oncogene) I think the survival information should also be in the .RData file.How can i know what all information is available in the give file. Please let me know any commands that show what type of information is available in the given file from a bio base package. Thank You rama kanth Download prohibited? No problem! To chat from any browser without download, Click Here: http://in.messenger.yahoo.com/webmessengerpromo.php __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Textpad help
I use TextPad for editting but have not tried to use it to submit code. I
use tinn-r which does the highlighting of the syntax and does submit to R
without any problems.
I will take a look at the link and see what happening when I try to use
TextPad for submission.
On 6/11/07, Mary Royerr [EMAIL PROTECTED] wrote:
Thx for the response John and Jim.
I had heard before that textpad is a good editor for R codes and one can
run
the R code from textpad itself and once we do that, the output is
displayed
on the textpad itself. The reason I wanted to use textpad is to make use
of
the syntax highlighting a la SAS.
So I installed textpad and according to the instructions given in
http://www.data-for-all.com/blog/?p=68 . everything worked out fine
until
I ran the following test code:
*setwd('C:/Temp/SampleR')
#
N-250
x-1:N
y-0.53*x +rnorm(x, 0.42*N, 0.14*N)
model1-lm(y~x)
print(summary(model1)) *
Once I submitted the R code using the SUBMIT R FILE in TOOLS from the
toolbar menu the following message displayed.
The filename, directory name, or volume label syntax is incorrect.
Tool completed with exit code 1
Now even If I run a simple code like *a-10;a* still the same message is
displayed.
I am kind of sure it is nothing wrong with the code but something is wrong
with the way the textpad installation. Any help? I am not a techie person.
Thx for your help
On 6/11/07, Mary Royerr [EMAIL PROTECTED] wrote:
I have installed textpad and tried running R code. But it gives me the
following error message.
The filename, directory name, or volume label syntax is incorrect.
Tool completed with exit code 1
Can you provide any help? I am not a technical person. So the help in
detail will be appreciated.
Thx
[[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
and provide commented, minimal, self-contained, reproducible code.
--
Jim Holtman
Cincinnati, OH
+1 513 646 9390
What is the problem you are trying to solve?
[[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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Textpad help
I looked at the link and it would never work for me since with TextPad it
starts a new session with each submission. With tinn-R it just sends it to
the current GUI session which makes debugging much easier since if an error
occurs you have the GUI to see exactly what was sent. I would strongly
suggest that you use tinn-r since the TextPad way does not allow incremental
development. Many of my scripts run for minutes to get to the point where I
want to test out some new code. So I am not even going to try it since if
something does error out, you have no way of seeing what else in the
environment might be causing the problem.
On 6/11/07, Mary Royerr [EMAIL PROTECTED] wrote:
Thx for the response John and Jim.
I had heard before that textpad is a good editor for R codes and one can
run
the R code from textpad itself and once we do that, the output is
displayed
on the textpad itself. The reason I wanted to use textpad is to make use
of
the syntax highlighting a la SAS.
So I installed textpad and according to the instructions given in
http://www.data-for-all.com/blog/?p=68 . everything worked out fine
until
I ran the following test code:
*setwd('C:/Temp/SampleR')
#
N-250
x-1:N
y-0.53*x +rnorm(x, 0.42*N, 0.14*N)
model1-lm(y~x)
print(summary(model1)) *
Once I submitted the R code using the SUBMIT R FILE in TOOLS from the
toolbar menu the following message displayed.
The filename, directory name, or volume label syntax is incorrect.
Tool completed with exit code 1
Now even If I run a simple code like *a-10;a* still the same message is
displayed.
I am kind of sure it is nothing wrong with the code but something is wrong
with the way the textpad installation. Any help? I am not a techie person.
Thx for your help
On 6/11/07, Mary Royerr [EMAIL PROTECTED] wrote:
I have installed textpad and tried running R code. But it gives me the
following error message.
The filename, directory name, or volume label syntax is incorrect.
Tool completed with exit code 1
Can you provide any help? I am not a technical person. So the help in
detail will be appreciated.
Thx
[[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
and provide commented, minimal, self-contained, reproducible code.
--
Jim Holtman
Cincinnati, OH
+1 513 646 9390
What is the problem you are trying to solve?
[[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
and provide commented, minimal, self-contained, reproducible code.
Re: [R] Rounding?
On Mon, 11 Jun 2007, jim holtman wrote:
your number 6.6501 is to large to fit in a floating point
number. It takes 56 bits and there are only 54 in a real number so the
system see it as 6.65 and does the rounding to an even digit; 6.6
I'll take it you mean a IEC60559 double, which has 53 bits in its implied
mantissa (it stores 52 and for normalized numbers the leading bit is 1
and not stored).
6.651 does fit into a real number (takes 54 bits) and this will
now round to 6.7
All you can say quickly is that its representation is greater than
6.65:
6.651 - 6.65
[1] 8.881784e-16
But I don't think that is the explanation. Remember that you are using
binary arithmetic, so each of these numbers is stored with representation
error. As the exact number stored is not '6.65', round-to-even does not
strictly apply.
I get
formatC(6.65,format=f,digits=1)
[1] 6.7
print(6.65, digits=2)
[1] 6.7
print(66.5, digits=1) # round to even really does apply
[1] 66
print(67.5, digits=1)
[1] 68
on (several of) my non-Windows systems, so I think this is a Windows
quirk. Remember the parsing and printing software has also to run in
binary on a limited-precision machine, and it does tend to be less
accurate on Windows than on other ix86 OSes. (Windows software also tends
not to implement round-to-even rules.)
On 6/11/07, BXC (Bendix Carstensen) [EMAIL PROTECTED] wrote:
I was a bit puzzed by:
formatC(6.65,format=f,digits=1)
[1] 6.6
So I experimented and found:
formatC(6.6501,format=f,digits=1)
[1] 6.6
formatC(6.651,format=f,digits=1)
[1] 6.7
round(6.6501,1)
[1] 6.7
round(6.651,1)
[1] 6.7
version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status
major 2
minor 5.0
year 2007
month 04
day23
svn rev41293
language R
version.string R version 2.5.0 (2007-04-23)
My machine runs Windows NT.
Is this intended or just a Windows facility?
__
Bendix Carstensen
Senior Statistician
Steno Diabetes Center
Niels Steensens Vej 2-4
DK-2820 Gentofte
Denmark
+45 44 43 87 38 (direct)
+45 30 75 87 38 (mobile)
+45 44 43 73 13 (fax)
[EMAIL PROTECTED] http://www.biostat.ku.dk/~bxc
This e-mail (including any attachments) is intended for the ...{{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
and provide commented, minimal, self-contained, reproducible code.
--
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
and provide commented, minimal, self-contained, reproducible code.
