[R] request for Unsubscribe

Dear Please unsubscribe my email from mailing list of R-Help. Thanks Amir Kavousi [EMAIL PROTECTED] - [[elided Yahoo spam]] [[alternative HTML version deleted]] __ R-help@r-proj

Re: [R] Using image function to plot character values

Eric Bell ericjbell.com> writes: > A secondary problem is that after I do the conversion, I need to know > what codes were converted to what numeric values. > Column-wise factor conversion can be handy, but often it bites. # Assume you have the data file m1 m2 A B B C C D #--- # m

Re: [R] request for Unsubscribe

Amir Kavousi wrote: > Dear > > Please unsubscribe my email from mailing list of R-Help. > > Please use the unsubscription mechanism at > https://stat.ethz.ch/mailman/listinfo/r-help > (near the bottom) -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_

[R] Constructing dummy variables for months for a time series object

I have a TS of monthly observations. head(data4) 1991(1) 1991(2) 1991(3) 1991(4) 1991(5) 1991(6) 12.00864 11.94203 11.98386 12.01900 12.19226 12.15488 Now I want to make 11 dummy variables indicating months. Therefore I did followings : For Jan : rep(c(rep(0,0), 1, rep(0,

[R] How to overlap two density plots?

Hi, How can I overlap two density plots? A <- c(8,10,10,11,11,11,12,12,12,12,11,11,11,10,10,7) B <- c(11,13,13,14,14,14,15,15,15,15,14,14,14,13,13,10) plot(density(A)) plot(density(B)) Regards, Emre P.S. There's a overlap.density function in package DAAG Is there another way to do th

Re: [R] force glm estimates to be nonnegative

You could use the BRugs package and formulate your prior knowledge about slopes in terms of a probability distribution. Michele Christina Itten kirjoitti: > Is there a way to force certain formula parameters to be nonnegative? > > What I want to do is to estimate student capacity over time, na

Re: [R] make check errors with R 2.7.0

On Fri, 25-Apr-2008 at 06:12AM +0100, Prof Brian Ripley wrote: > Note that the problem is not that you got the message, but an apparent > whitespace issue in the output. That's clearer in this version: note that > the output for reg-IO.Rout.save is shifted left one column. But I don't > trust

Re: [R] Constructing dummy variables for months for a time series object

> GenDummyVar <- function (NumMonth) rep(c(rep(0, NumMonth-1), 1, rep(0, 12-NumMonth)), 17) > NameDummy <- c("DJan", "DFeb", "DMar", "DApr", "DMay", "DJun", "DJul", "DAug", "DSep", "DOct", "DNov", "DDec") > for (NumMonth in 1:12) assign(NameDummy, GenDummyVar (NumMonth)) > DJan [1] 0 0 0 0 0 0 0

Re: [R] Constructing dummy variables for months for a time series object

Oops.. the last line should have been : for (NumMonth in 1:12) assign(NameDummy[NumMonth], GenDummyVar (NumMonth)) On Fri, Apr 25, 2008 at 10:18 AM, N. Lapidus <[EMAIL PROTECTED]> wrote: > > GenDummyVar <- function (NumMonth) rep(c(rep(0, NumMonth-1), 1, rep(0, > 12-NumMonth)), 17) > > NameDummy

Re: [R] request for Unsubscribe

Amir Kavousi schrieb: > Dear > > Please unsubscribe my email from mailing list of R-Help. > > Thanks > > Amir Kavousi > > [EMAIL PROTECTED] > you can unsubsribe at the following page by yourself. (including password reminder) https://stat.ethz.ch/mailman/listinfo/r-h

[R] Non-linear system of equations

Hello R users, I am trying to estimate the parameters of a bimodal normal distribution using moments matching, so I have to solve a non-linear system of equations. How can I solve the following simple example? x^2 - y^2 = 6 x – y = 3 I heard about nlsystemfit, but I don’t know how to run it e

Re: [R] Using image function to plot character values

Eric Bell wrote: > Hello. I am a newbie to R. If I should be reading some FAQ or manual > that could help answer my question please tell me and I will go there. > > Problem: > > I have a spreadsheet that contains a character code in each cell. The > columns in the spreadsheet represent time and

[R] grid() : start values?

Hi, is not there any way to tell grid() the starting value? With nx,ny I get the right number of grid cells, but not on the exact values I would like them. (I know I can do it with abline(), just asking for an easier way, i.e. grid(nx,nx,start.x,start.y)) An alternative way (which would solve anot

Re: [R] Non-linear system of equations

2008/4/25 Radka Pancheva <[EMAIL PROTECTED]>: > I am trying to estimate the parameters of a bimodal normal distribution > using moments matching, so I have to solve a non-linear system of equations. > How can I solve the following simple example? > > x^2 - y^2 = 6 > x – y = 3 > > I heard abou

Re: [R] How to overlap two density plots?

"Emre Unal" <[EMAIL PROTECTED]> wrote in news:[EMAIL PROTECTED]: > Hi, > > How can I overlap two density plots? > > > A <- c(8,10,10,11,11,11,12,12,12,12,11,11,11,10,10,7) > B <- c(11,13,13,14,14,14,15,15,15,15,14,14,14,13,13,10) > > plot(density(A)) Instead: lines(density(B)) -- David Win

Re: [R] Non-linear system of equations

Hello Paul, Thank you for your quick answer. I have tried to use your advice and to estimate the parameters of beta distribution with moments matching. This is my code: ex <- 0.3914877 ex2 <- 0.2671597 my.mm <- function(x){ p <- x[1] q <- x[2] p <- .Machine$double.eps

Re: [R] grid() : start values?

On 4/25/2008 5:32 AM, Agustin Lobo wrote: > Hi, > is not there any way to tell grid() the starting value? With nx,ny > I get the right number of grid cells, but not on the exact > values I would like them. (I know I can do it with abline(), just > asking for an easier way, i.e. grid(nx,nx,start.x,s

Re: [R] How to overlap two density plots?

On Fri, Apr 25, 2008 at 10:57:27AM +0300, Emre Unal wrote: > > How can I overlap two density plots? > > A <- c(8,10,10,11,11,11,12,12,12,12,11,11,11,10,10,7) > B <- c(11,13,13,14,14,14,15,15,15,15,14,14,14,13,13,10) plot(density(A)) lines(density(B)) cu Philipp -- Dr. Philipp Pagel L

[R] No rule to make target 'R/Copy'

Dear all, I am currently building a package with C code (in scr folder). Everything was ok, but now, I get the following error message when compiling the package using R CMD INSTALL: make[2]: *** No rule to make target 'R/Copy', needed in [...] What does that mean? I can still compile other pack

Re: [R] Use of survreg.distributions

--begin included message --- I am using survreg(Surv()) for fitting a Tobit model of left-censored longitudinal data. For logarithmic transformation of y data, I am trying use survreg.distributions in the following way: tfit=survreg(Surv(y, y>=-5, type="left")~x + cluster(id), dist="gaussian",

Re: [R] Non-linear system of equations

> x^2 - y^2 = 6 > x – y = 3 > > You can also try this # function f <- function(x) { y <- numeric(2) y[1] <- x[1]^2-x[2]^2-6 y[2] <- x[1]-x[2]-3 y } # function values transformed to scalar # minimising fnorm this way is not the best method of finding a solution for

Re: [R] Non-linear system of equations

Try to change the initial values of the parameters with, for instance, p0 <- rnorm(2) But you sure that your system has a solution, Evgeniq? Paul 2008/4/25 Radka Pancheva <[EMAIL PROTECTED]>: > Hello Paul, > > Thank you for your quick answer. I have tried to use your advice and to > estimate

Re: [R] No rule to make target 'R/Copy'

Do you have a space in some file name? Starting 'Copy '? On Fri, 25 Apr 2008, Ardia David wrote: > Dear all, > I am currently building a package with C code (in scr folder). > Everything was ok, but now, I get the following error message when > compiling the package using R CMD INSTALL: > make[

Re: [R] R and condor

Paul, This is what some of our condor files look like. The script actually goes through and cleans out our log directory (hence the name): --- CleanCondorLogs.condor --- JobName = CleanCondorLogs LogHeader = $(JobName).$(Cluster).$(Process).$$(Machine) input = /CleanCondorL

Re: [R] R Newbie Question/Data Model

[EMAIL PROTECTED] wrote: > Given a data set and a set of predictors and a response in the data, > we would like to find a model that fits the data set best. > Suppose that we do not know what kind of model (linear, polynomial > regression,... ) might be good, we are wondering if there is R-package(

[R] package compilation: including external libraries in windows compilation

Dear R package developers: I need help on getting a package compiled for windows. I am trying to use definitions and routines from a set of libraries in that are in LINUX under #include #include #include #include but these are unavailable in windows. I found that we can get these by including

Re: [R] Non-linear system of equations

Clearly, a family of solutions is (p,q) = (0,k), with k != 0. Paul On Fri, Apr 25, 2008 at 1:58 PM, Paul Smith <[EMAIL PROTECTED]> wrote: > Try to change the initial values of the parameters with, for instance, > > p0 <- rnorm(2) > > But you sure that your system has a solution, Evgeniq? > >

Re: [R] Non-linear system of equations

Hi Radka, The problem lies in your function my.mm(). Especially, the following lines: p <- .Machine$double.eps q <- .Machine$double.eps Even if you take out these 2 lines, still the definition is not correct, I think, since there is a trivial answer x* = c(0, q). I am not sure how you int

Re: [R] Non-linear system of equations

Radka, Generally, there is no need to increase `maxit' to beyond default. If you are unable to obtain a reasonable approximation to your system within `maxit', then you should do the following (not necessarily in that order): 1. Try a few different starting values and see if you can obtain a re

[R] Extracting coefficients' standard errors from linear model

Dear all, Hope this question is not too trivial... In order to correct standard errors from an estimation of a fixed effects regression model y need to extract the vector of standard errors of the coefficients of a simple linear model estimated using lm(var1~var2+var3+var4) Unfortunately I ca

Re: [R] Non-linear system of equations

Try this: > library(Ryacas) > x <- Sym("x") > y <- Sym("y") > Solve(List(x^2+y^2==6, x-y==3), List(x,y)) expression(list(list(x == root(6 - y^2, 2), y == y))) 2008/4/25 Radka Pancheva <[EMAIL PROTECTED]>: > Hello R users, > > I am trying to estimate the parameters of a bimodal normal distributio

[R] remarks on function pnbeta

Dear Developers, The pnbeta function has been reviewed recently in the article of A. Baharev, S. Kemény, On the computation of the noncentral F and noncentral beta distribution, Statistics and Computing, 2008, in press. Preprint of the paper is available here: http://reliablecomputing.eu/publicat

Re: [R] package compilation: including external libraries in windows compilation

The standard C define is WIN32, not Win32 (which R uses as private define). You could have PKG_CPPFLAGS=-DWin32 in Makevars.win, but this is unnecessary. See e.g. Rmath.h. On Fri, 25 Apr 2008, Jonathan Lees wrote: > Dear R package developers: > > I need help on getting a package compiled for

Re: [R] Extracting coefficients' standard errors from linear model

Uli Kleinwechter <[EMAIL PROTECTED]> wrote in news:[EMAIL PROTECTED]: > Dear all, > > Hope this question is not too trivial... > > In order to correct standard errors from an estimation of a fixed > effects regression model y need to extract the vector of standard > errors of the coefficients

Re: [R] Documentation General Comments

I realize the R developers are probably overwhelmed and have little time for this, but the documentation really needs some serious reorganizaton. My reply: I'm quite new to R and so have spent a lot of time in the last few days reading documentation both online and text. I'm incredibly impres

[R] join chars in loops

Hi, Is it possible to create new vars in a loop For instance: for (i in 1:11) var'i' <- assign something output = var1 var2 ... var11 Is there a way to combine/join the value of 'i' to the string/char var in de loop. -- View this message in context: http://www.nabble.com/join-chars-in-lo

Re: [R] join chars in loops

for (i in 1:11){ assign(paste('var', i, sep=''), value) } On Fri, Apr 25, 2008 at 10:40 AM, Dirkheld <[EMAIL PROTECTED]> wrote: > > Hi, > > Is it possible to create new vars in a loop > > For instance: > > for (i in 1:11) > var'i' <- assign something > > output = > var1 > var2 > ... > var11 >

Re: [R] Extracting coefficients' standard errors from linear model

Try: summary(lm.D9)[["coefficients"]][,2] On Fri, Apr 25, 2008 at 10:55 AM, Uli Kleinwechter < [EMAIL PROTECTED]> wrote: > Dear all, > > Hope this question is not too trivial... > > In order to correct standard errors from an estimation of a fixed > effects regression model y need to extract th

Re: [R] join chars in loops

look at ?assign(), e.g., for (i in 1:11) { assign(paste("var", i, sep = ""), i) } var1 var2 var10 but I'd prefer to use a list. I hope it helps. Best, Dimitris Dimitris Rizopoulos Biostatistical Centre School of Public Health Catholic University of Leuven Address: Kapucijnenvoer 35, L

Re: [R] matrix from list

You may want to read the help page for '[' and '[[' a few more times (the differencese can be subtle). The '[[' only returns a single element from a data structure (but in the case of a list, that single element can be quite complex). So something like list[[1:2]] does not make sense because you

Re: [R] No rule to make target 'R/Copy'

No... don't think so. I rerun my computer, moved some folders (I deleted the folder with .C file), rerun the compilation, and it worked... Not sure to understand why. Any idea? Prof Brian Ripley wrote: > Do you have a space in some file name? Starting 'Copy '? > > On Fri, 25 Apr 2008, Ardia Da

Re: [R] Extracting coefficients' standard errors from linear model

Or use: mod <- lm(var1 ~ var2 + var3 + var4) coef(summary(mod))[, 2] or coef(summary(lm.D9))[, "Std. Error"] if you prefer using column names. coef() extracts the model coefficients from the lm object and the additional content in a summary.lm object. There are accessor functions f

Re: [R] Sum of random values

You might want to look at the distr package (and its relatives). It provides methods for calculating the distribution function of combinations (the sum is one) of other distributions. I'm not sure how you would convert your percentiles to a distribution function, but there may be a way in the doc

Re: [R] join chars in loops

... That is, use the construction thelist <- lapply(1:11,function(i){...return something}) I would go further then Dmitris and say that this should be the canonical answer to this frequently asked question and that perhaps in 7.21 in the R FAQ (Version 2.6.2 -- I haven't yet upgraded) -- "How ca

[R] fix variance parameter values for lmer estimation

Dear list-members, a model is fit with lmer, but I want to force the variance parameter values to be as defined by me I thought, use 'start' to specify initial values and only allow for one iteration ? my question is how to do that ? to specify the 2x2 matrix of variance parameter values: start=

[R] FASeg

Hi All, I was using FASeg package. Now when I am trying to install this library I get the message Warning message: package 'FASeg' is not available Can someone please suggest something. Thank you vidhu [[alternative HTML version deleted]] __ R

[R] Rows assignment

Hi to all members, i encountered a weird problem when trying to assign subset of rows in a matrix. Simple example works fine: z<-matrix(NA,ncol=3,nrow=10) z[,1]<-1:10 y<-matrix(1:30,ncol=3,nrow=10) y[,1]<-5:14 z[z[,1]%in%y[,1],2]<-y[y[,1]%in%z[,1],2] z z[z[,1]%in%y[,1],]<-y[y[,1]%in%z[,1],] z W

Re: [R] fix variance parameter values for lmer estimation

On 4/25/08, Wilfried Cools <[EMAIL PROTECTED]> wrote: > Dear list-members, > a model is fit with lmer, but I want to force the variance parameter values > to be as defined by me > I thought, use 'start' to specify initial values and only allow for one > iteration ? > my question is how to do

Re: [R] Use of survreg.distributions

Hello Dr. Therneau:   Thank you for your response. Let me explain to you want I want. My y variable(time) is normal and would like to fit the model on logarithmic transformation of y (log(y)). I tried to run codes according to your suggestion: tfit=survreg(Surv(y, y>=-5, type="left")~x + cluster(

[R] making a dll for windows eviornment for R

Hi All, I have a .c file and I need to make a dll so that it could be called by R. I have successfully made .so from .c file and its working great. But i have no clue of making dll. I tried googling for help ...but not much of use Please suggest something regards Vidhu [[alternative HTM

Re: [R] Documentation General Comments

Each pdf reference manual includes an index of functions classified according to keywords. I wonder why indexes in on-line documentation do not include similar categorisation. It would be much easier to figure out all related functions and topics. Thanks, Vitalie. __

Re: [R] making a dll for windows eviornment for R

On 4/25/2008 1:12 PM, Vidhu Choudhary wrote: > Hi All, > I have a .c file and I need to make a dll so that it could be called by R. > I have successfully made .so from .c file and its working great. But i have > no clue of making dll. I tried googling for help ...but not much of use > > Please su

Re: [R] Obtaining correlation parameters for multiple rows

Hi Jorge, That was great! Thank you very much for the suggestion. I just had to make a couple of minor changes (e.g., pearson vs. spearman) and it worked perfectly. Thanks, Joao Pedro Hi, I'm sure it could be better but try this: # F statistics based on lm FSTAT=function(y,x) summary(lm(y~

Re: [R] Differentiate alphanumeric vs numeric strings

This is fantastic. As a future note to myself and to others who may stumble upon this: further documentation for all one's heart could desire is in the base help system at. Regular Expressions as used in R - regexp Pattern Matching and Replacement - grep Split the Elements of a Character Vector

[R] hist densities

This is a reply to Luke Spadavecchia's question about hist densities. If you want to construct a histogram where the bar heights are the percentage of observations in the intervals you can use this code: x<-rnorm(n=5000,mean=1.5,sd=0.25) h<-hist(x) h$density<-h$counts/sum(h$counts) plot(

Re: [R] Constructing dummy variables for months for a time series object

The simplest way is probably to generate them as the columns of a matrix. Let's suppose you have data from Jan 1991 up to Mar 2008. > X <- outer(rep(month.abb, length = 16*12+3), month.abb, "==") + 0 > ## the "+ 0" will convert logical to numeric > dim(X) [1] 195 12 > dimnames(X)[[2]] <- mont

[R] Equivalent of Excel pivot tables in R

Can somebody tell me how to do the equivalent of a pivot table in R ? For example, if I have : var1var2 var3 a x10 b y20 a z10 b z20 a z10 b z20 I could have : x y z a

Re: [R] Equivalent of Excel pivot tables in R

Hi, Try this: x="var1var2 var3 a x10 b y20 a z10 b z20 a z10 b z20" yourdata=read.table(textConnection(x),header=TRUE) attach(yourdata) res=tapply(var3,yourdata[,-3],sum) # With tapply! res[is

Re: [R] Equivalent of Excel pivot tables in R

See Hadley's reshape package http://had.co.nz/reshape/ and on CRAN. On Fri, Apr 25, 2008 at 2:54 PM, ppaarrkk <[EMAIL PROTECTED]> wrote: > > Can somebody tell me how to do the equivalent of a pivot table in R ? > > > For example, if I have : > > var1var2 var3 > a x10 > b

Re: [R] Equivalent of Excel pivot tables in R

Try: with(x, tapply(var3, list(var1, var2), length)) with(x, tapply(var3, list(var1, var2), sum)) Or with xtabs xtabs(as.logical(var3) ~ var1 + var2, data = x) xtabs(var3 ~ var1 + var2, data = x) On 4/25/08, ppaarrkk <[EMAIL PROTECTED]> wrote: > > > Can somebody tell me how to do the equiv

Re: [R] matrix from list

Greg Snow wrote: > The '[[' only returns a single element from a data structure I know but that is precisely what I find arbitrary. Anyway you are right that it would still return the kind of object, only subsetted, which is not I want. As someone kindly pointed out to me offline, besides the u

[R] Ubuntu vs Mac OS X

Hi ! I have installed R-2.7.0 64-bits on a computer where the new Ubuntu version (8.04) was installed. This computer is a 3.2 GHz Dual-core with 2 Gb of RAM. To test how fast (I though !) this machine is, I compared two simple matrix manipulations between Ubuntu and my laptop (a MacBook Pro 2.

Re: [R] Equivalent of Excel pivot tables in R

Hi, I forgot the first part (I'm sorry): table(var1,var2) var2 var1 x y z a 1 0 2 b 0 1 2 HTH, Jorge On Fri, Apr 25, 2008 at 7:25 PM, Jorge Ivan Velez <[EMAIL PROTECTED]> wrote: > > Hi, > > Try this: > > x="var1var2 var3 > a x10 > b y20 > a

Re: [R] Constructing dummy variables for months for a time series object

First generate a factor and then calculate its model matrix: month. <- factor(cycle(x), labels = month.abb) model.matrix(~ month.)[,-1] Actually you may not even need the dummy variables at all if your purpose is to do a regression. You can use the factor directly: lm(x ~ month.) By the way, b

Re: [R] Ubuntu vs Mac OS X

On Apr 25, 2008, at 4:34 PM, Guillaume Blanchet wrote: > Hi ! > > I have installed R-2.7.0 64-bits on a computer where the new Ubuntu > version (8.04) was installed. This computer is a 3.2 GHz Dual-core > with > 2 Gb of RAM. > > To test how fast (I though !) this machine is, I compared two simp

[R] R question about prediction from a fitted model

Hi, I have a question about predicting new values from a fitted model in R. For example, if i fit a linear model fit<-lm(y~x) predict(fit) will give the fitted values for each x. I am wondering if there is a way to do some prediction of the fitted for some new values of x that are not in the d

Re: [R] R question about prediction from a fitted model

Hi Lisa, See examples in ?predict.lm HTH, Jorge On Fri, Apr 25, 2008 at 7:47 PM, Lisa <[EMAIL PROTECTED]> wrote: > Hi, I have a question about predicting new values from a fitted model in R. > > For example, if i fit a linear model > > fit<-lm(y~x) > > predict(fit) will give the fitted value

[R] Summarize data for MCA (FactoMineR)

Hi :-) I'm new to R and I started using it for a project (I'm the CS guy in a group of statisticians helping them find out how to solve issues as they come out). This is my first post to the list and I am starting to learn R. Well, they were used to doing MCA analysis in other programs where the

[R] Details regarding the nnet package

Hello, I've recently been using the nnet package to do some basic forecast predictions. I've found the package to be quite useful and I am getting some good results. However, I am in the midst of writing a small paper on the results I am getting and wish to clarify some things about the nn

[R] function clogit

Hello, I am using the clogit (conditional logistic regression) on a simple data set which is not related to survivorship, which I understand to be fine. I have trouble understanding the output. I would like to find parameter estimates of the logistic models I am constructing with R, and am u

[R] 3 questions: debug R script, multi-level sorts, and multi-gsub

Hi, 1) I ran a script say test.R. It ran and terminated with an error. However, I am not sure where the error occur. > source("test.R") Loading required package: tools Welcome to Bioconductor Vignettes contain introductory material. To view, type 'openVignette()'. To cite Bioconductor, see

Re: [R] 3 questions: debug R script, multi-level sorts, and multi-gsub

When it stops with the error you can issue a traceback() command or run the script with echo=TRUE: source("test.R", echo = TRUE, max = 1000) On Fri, Apr 25, 2008 at 11:28 PM, Ng Stanley <[EMAIL PROTECTED]> wrote: > Hi, > > 1) I ran a script say test.R. It ran and terminated with an error. Howe

Re: [R] R question about prediction from a fitted model

Hi Jorge, thank you very much for your reply. I checked and have got the prediction of new values for lm or glm or nls, the new question is, is there any prediction for the cox model? I tried predict(Surv(y~x),newdata), it only gave the fitted values. Any clue on this? thanks a lot! On 4/25/08,