[R] block matrices

Dear all, I do not know how to deal with block matrices in R. For example I have 3 matrices A, B and C. And I want to produce a new matrix of this form ( A B 0 ) ( 0 0 C ) where A, B and C are one-row matrices. Apart from A, B and C, all the coefficients are 0. Is there an easy solution in R

Re: [R] block matrices

t; On Wed, Feb 24, 2010 at 12:56 PM, Gustave Lefou > wrote: > > Dear all, > > > > I do not know how to deal with block matrices in R. > > > > For example I have 3 matrices A, B and C. > > > > And I want to produce a new matrix of this form > >

[R] Emacs for R

Dear all, >From the recent discussion, I have wondered where I could find some quick step documentation on Emacs for R (especially on Windows). All I have found is that 80 pages pdf http://ess.r-project.org/ess.pdf Maybe I am asking for too much ? Best, Gustave [[alternative HTML versi

Re: [R] Emacs for R

.., DSC 2001 Proceedings of the > 2nd International Workshop on Distributed Statistical Computing, March > 15-17, Vienna, Austria > > Cheers > > David Cross > d.cr...@tcu.edu > www.davidcross.us > > > > > > > On Mar 1, 2010, at 10:41 AM, Gustave Lefou wrot

[R] lines - graphics

Dear R users, I have two vectors : x and y. x takes negative values and y positive ones. I write : > plot(x) > lines(y) However because y takes positive values, I cannot see y. What can I do ? Thank you very much, Gustave [[alternative HTML version deleted]]

Re: [R] lines - graphics

Thanks Mario and Jim 2010/4/13 jim holtman > plot(x, ylim=range(c(x,y)), xlim=c(0,max(c(length(x),length(y > > might work. > > On Tue, Apr 13, 2010 at 7:19 AM, Gustave Lefou wrote: > >> Dear R users, >> >> I have two vectors : x and y. x takes negative

[R] return of a function

Dear R users, I have a function which takes as arguments big arrays, say : w, x , y and z. My function changes these arrays and I want them as result/output. I have tried to write return(w,x,y,z), and thus to replace the previous w, x, y and z. It does not seem to work. What can I do ? Thank y

Re: [R] return of a function

Thank you David, Jim and Gunter. Actually I knew about the list thing. Everything got confused in my head :-( I guess I am a bit tired... 2010/4/16 David Winsemius > > On Apr 16, 2010, at 12:02 PM, Gustave Lefou wrote: > > Dear R users, >> >> I have a function whic

[R] warning message cran

Dear R users, I have downloaded some packages. The beginning is very slow until I am allowed to choose a CRAN server. Then it is faster. I get a warning message each time at the end of the download : "The downloaded packages are in C:\etc Warning message: In open.connection(con, "r") :

Re: [R] warning message cran

Thank you very much Siri and Uwe. So there is no problem with the packages downloaded. 2010/4/27 Uwe Ligges > > > On 27.04.2010 12:21, Gustave Lefou wrote: > >> Dear R users, >> >> I have downloaded some packages. The beginning is very slow until I am >> a

Re: [R] timing a function

Dear all, I have a function f(x). > ptm <- proc.time() > y=f(x) > proc.time() - ptm works fine for me. However it is not possible to write > system.time(y=f(x)) and > system.time(f(x)) does not store the value of f(x). Should I prefer the 3 lines with proc.time ? Thank you very much, Gusta

Re: [R] timing a function

Thank you Bill, Peter and Barry. 2010/5/17 Barry Rowlingson > On Mon, May 17, 2010 at 6:24 PM, Peter Ehlers wrote: > > > Try > > system.time(y <- f(x)) > > > > and see ?"=". > > > > -Peter Ehlers > > Ah ha. That explains the curly brackets I saw in a posting with > system.time on stack ove

Re: [R] timing a function

Thank you Alexander 2010/5/17 Alexander Shenkin > You could also put the call to system.time inside the function itself: > > f = function(x) { >system.time({ >... #function's code >ret_val = ... >}); flush.console(); >return ret_val; > } > > i s'pose you'd miss out on

Re: [R] timing a function

Dear all, Just one last question. There seems to be no problem in writing > z = system.time(y <- f(x)) or > z <- system.time(y <- f(x)) Then z contains the named vector of the elapsed times, and y the value of the function f(x). Am I right ? Thank you very much, Gustave 2010/5/17 Alexander

[R] fitdistr errors

Hello, I have got the following error message (translated from French to English) using "fitdistr" : " Error in fitdistr(nira, "weibull") : optimization failed Furthermore : There are 50 warnings or more (use warnings() to see the first 50) " I used "fitdistr" in a loop and I think the loop stop

[R] substitute in graphics

Hello, I have to do a few graphics of the same function and this function is parametrized by two arguments. What I would like is to be able to change the value of these two arguments without changing the plot command. So as to copy paste. I tried the following : x=1:100 eta=10 beta=5 plot(x,h(x

Re: [R] substitute in graphics - Uwe's help desk

st(myeta=eta,mybeta=beta)) ) Any idea ? Thank you very much 2008/5/9 Henrique Dallazuanna <[EMAIL PROTECTED]>: > Try this: > > plot(x, log(x), > xlab = "x", ylab = "h(x)", > main = bquote(Failure~rate~from~W(eta == .(eta), beta == .(beta))) > &

Re: [R] substitute in graphics - Uwe's help desk

* Sigma[x] * "=" * bgroup("(",atop(s1~~s3,s2~~s4),")"),param.list)) So like in my problem the problem seems to be ==. Thank you very much 2008/5/10 Gustave Lefou <[EMAIL PROTECTED]>: > Thank to both of you. > > I found an interesting document by Uwe

[R] box-constrained

Hello everybody, I have a question about box-constrained optimization. I've done some research and I found that optim could do that. Are there other ways in R ? Is the following correct if I have a function f of two parameters belonging for example to [0,1] and [0,Infinity] ? optim(par=param, fn=

Re: [R] box-constrained

ins 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:

Re: [R] box-constrained

Hello everybody, I have another question. I have seen there is a function called "constrOptim" in R. Is it better than "optim", for example to optimize a function f of two parameters belonging to [0,1] and [0,Infinity] ? Do the methods supplied like Nelder-Mead are better than those of optim ?

Re: [R] box-constrained

whether constrOptim is useful for boundary problems, or if optim is sufficient for box-constrained optimization. Thank you. 2008/3/10, Paul Smith <[EMAIL PROTECTED]>: > > On Sun, Mar 9, 2008 at 9:10 PM, Gustave Lefou <[EMAIL PROTECTED]> > wrote: > > I have another que

[R] Multi arguments - return

Hello, I wanted a function to have two vectors x,y as its output. I wrote return(x,y) inside its definition. It worked but I got a warning : it is obsolete to do like that. So I tried to put x,y in a list(x,y) and to return(list(x,y)). It worked. I called the return of the function result. But re

Re: [R] Multi arguments - return

Thank you for that quick responses 2008/3/11, [EMAIL PROTECTED] <[EMAIL PROTECTED]>: > > >From: Gustave Lefou <[EMAIL PROTECTED]> > >Date: 2008/03/11 Tue PM 12:32:25 CDT > >To: [EMAIL PROTECTED] > >Subject: [R] Multi arguments - return > > if you'

Re: [R] Multi arguments - return

Yep ! It works perfectly. 2008/3/11, [EMAIL PROTECTED] <[EMAIL PROTECTED]>: > > >From: Gustave Lefou <[EMAIL PROTECTED]> > > >Date: 2008/03/11 Tue PM 12:57:59 CDT > >To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>, > Phil Spector <[EMA

[R] set the language

Hello, I would like to change my R language from German to French or English. I have read messages in the archives and also R Installation and Administration paragraph 7.2 but I did not get it all. I have to set LANGUAGE=en (what would it be for French, "fr" ?), but I don't know where. I am unde