Re: [R] How to generate a smoothed surface for a three dimensional dataset?

Hi Jun, I'd for a local fitting by means of the loess() function. Assuming z is the third axis: x<-runif(25) y<-runif(25) z<-runif(25) xyz.fit <- loess(z~x+y,control=loess.control(surface='direct'),span=.5,degree=1) #tune parameters as you like z.predict <-matrix(predict(xyz.fit,cbind(x,y)),5,5

Re: [R] exponent function help??

if just integer exponentiation is enough, then I'd suggest: expf <- function(x, n) { return(prod(rep(x, abs(n)))^sign(n)) } cheers. /federico On Wed, Jul 13, 2011 at 6:24 AM, Ben Bolker wrote: > mousy0815 gmail.com> writes: > > > I'm trying to make a function that will output the exponent..

Re: [R] Syntax Highlightning and Editor for Linux

I have been trying emacs with ess, Kate, plain gedit..nothing really satisfied me.. but then I found a nice plugin for gedit: Rgedit ! it supports split screen with R terminal window (you can also personalize colour if, like me, do like to stick with old habits :) ), moreover gedit has a brackets

[R] DIF Analysis starting from a gpcm class object

such a thing by myself? and if so, how can I extract what I need from the model object? thanks to everybody /federico ----- Dr. Federico Andreis Università degli Studi di Milano-Bicocca, PhD Student MEB Department, Karolinska Institutet, Stockholm, Visit

Re: [R] Exporting NMDS distance matris to csv

On Tue, Jul 20, 2010 at 5:54 PM, Gavin Simpson wrote: > On Tue, 2010-07-20 at 08:24 -0400, Graves, Gregory wrote: > > If you submit these lines, you end up with variable "vare.dis". I want > > to export vare.dis to csv. Stuck I am. > > Hi Yoda ;-) > > I want THIS on fortune() ! ahhahah so

Re: [R] Multilevel IRT Modelling

Thanks Stuart, I already had some of those papers, will check the others! best Federico On Wed, Jul 14, 2010 at 9:05 PM, Stuart Luppescu wrote: > On Wed, 2010-07-14 at 04:31 -0700, Dr. Federico Andreis wrote: > > does anybody know of a package (working under Linux) for multileve

Re: [R] Want to exclude a column when using sample function

I guess you could just use as an argument to sample data[,-c(1,2)] where 1 and 2 are id and pID50 column number On Wed, Jul 14, 2010 at 7:17 PM, Addi Wei wrote: > > id pID50 apol a_acca_acid a_aro a_base a_count > 1 mol.11 3.63 -0.882267 -0.527967 -0.298197 -1.032380

[R] Multilevel IRT Modelling

Dear All, does anybody know of a package (working under Linux) for multilevel IRT modelling? I'd love to do this without having to go on WINSTEPS or the like.. thanks for the attention! Federico Andreis - Dr. Federico Andreis Università degli Studi di Milano-Bicocca, PhD Studen

[R] Likelihood Optimization With Categorical Variables

o you have any suggestion? Any other routine which could deal with categorical (along with non categorical) variables without for the aim of optimization? thanks in advance Federico Andreis Università degli Studi di Milano Bicocca, PhD Student MEB, Karolinska Institutet, Visiting PhD Student