similar to: supsmu vs. ppr

Dear R listers -- The program below does the following tasks: 1. It creates a file (wintemp4) that is a subset of alldata4 consisting of "winner" records in 50 industry groups (about 5400 obs); 2. It defines a function (myppr1) that runs the ppr function in modreg once to generate goodness of fit (sum of squared errors) measures by number of terms included in model and then reruns

Sorry about the lack of detail. I am running R v.1.2.2. I can recast my question (which I think I have partially answered) more succinctly as follows: 1. This seems to work (note that group takes values 1,2,3,4, or 5): my.newfun <- function(x) myfile <- lm(award ~ ilogemp + ilogage, x) test.by <- by(wintemp, as.factor(wintemp$group), my.newfun) 2. This does not work (leaving aside

Dear R listers -- I have several questions about using the ppr command in the modreg module. I discovered -- quite by accident -- that if I re-order the data, I obtain different results. The output below shows what I mean. I have two datasets (dataset1 and dataset2) that are identical (tested using proc compare in SAS) except for the fact that the records are in different order. Below I have

Dear R list: I want to make separate estimates for each level of the variable "group." After consulting many sources I am stumped as to why the following does not work: > wintemp <- subset(alltemp, winner==1) > my.ppr <- function(x) + { + if(nrow(x) >= 50) { + pprfile <- ppr(award~ilogemp, data=x,nterms=5,max.terms=10,optlevel=3) + summary(pprfile) +

Dear R users, I am looking for the source code of the implementation of ppr (Projection Pursuit Regression) in R. It will be great if citations of the source papers on which the implementation is based, are also provided. Thank you, Arvind Iyer, Grad student, Deptt. of Biomedical Engineering Viterbi School of Engineering University of Southern California, Los Angeles [[alternative HTML

supsmu(periodic=TRUE) can crash R by reading before start of array. To reproduce: set.seed(1) xx <- runif(29000) yy <- rnorm(29000) span <- 0.49 i <- 1 while(i < 200){ cat(i,"\n") int <- supsmu(xx,yy,periodic=T,span=span) i <-i+1 } results in: 1 2 3 4 5 6 7 8 9 Program received signal SIGSEGV,

Hi all, I am doing a projection pursuit regression using the ppr() function from modreg. I would also like to use predict.ppr(). However, I cannot find any information about it in the help files. There is a link to predict.ppr in the index for modreg, but that link is to the help for ppr(). Has predict.ppr() not been implemented? If not, does anyone have a suggestion as to how to implement

Hi, Recently, I installed the 1.1.0 version of R (for Windows), since it includes an implementation of Projection Pursuit (I failed to write my own version of PP as a standalone C++ program). As far as I know, R offers two interfaces/sintax for the ppr() function. The first one requieres a regression formula and a data frame. The other requieres X, a matrix with the explanatory variables, and Y,

Dear all, I have been doing tests using SVM, random forests and PPR. The data is from a data stream (that is, the data for training and for test is always increasing / changing). With SVM and random forests everything is ok, but with ppr there are situations where it crashes. For the examples I have used I noticed that if one of the variables has just one value (it can happen), it crashes for

Folks, The projection pursuit regression function in the base R seems to crash when the optimization level is set to zero, i.e. the initial ridge terms are accepted without refitting. I encountered this problem in an out-of-sample prediction exercise using predict. But further investigation suggests the issue is with the ppr fit and predict just sppeds up the crash. The other optlevels seem to be

The pursuit projection packages ppr is an excellent contribution to R. It is great for one-to-three ridge fits, often somewhat intuitive, and for multi-ridge fits, where it at least describes a lot of variance. Like many folk, I need to report the fits obtained from ppr to the greater, outside, non-R world. It is fairly obvious how to use the terms alpha and beta to report on

Full_Name: Hugh Miller Version: 2.8.1 OS: XP Submission from: (NULL) (128.250.24.101) Hi there, I've been looking at approaches that use the projection pursuit regression function fairly (ppr) heavily, and have discovered that it crashes my R system on occasion. It only happens with the inputs are pathological in some way I don't understand. I have pasted such an example below. Any

Dear list, When I use VennDiagram package, I got a error as follow: venn.plot <- draw.pairwise.venn( area1 = 3186, area2 = 325, cross.area = 5880); Error in draw.pairwise.venn(area1 = 3186, area2 = 325, cross.area = 588) : Impossible: cross section area too large. Does anyone have suggestion? Thank you.

I'm reading Torgo (2010) *Data Mining with R*<http://www.liaad.up.pt/~ltorgo/DataMiningWithR/code.html>in preparation for a class I'll be teaching next quarter. Here's an example that is very non-functional. > pH <- c(4.5,7,7.3,8.2,6.3) > names(pH) <- c('area1','area2','mud','dam','middle') > pH area1 area2 mud dam

When handed an argument with all NAs, supsmu() causes a sigfault causing termination of R. The following example reproduces the problem (on a linux gentoo system), but I have verified the same behavior on a solaris builds as well as several versions of R. x <- (1:100)/10; y <- sin(pi*x) + rnorm(length(x)); tmp <- supsmu(x,y); # works y[c(2,5,10)] <- NA; # a few NAs

Hi, I have a dataset with x and y coordinates and in each point I have an identity of point, in some cases I can have more then one identity by point. My dataset is something like this: > x <- rep(c(1:4),4) > y <- rep(c(1:4),c(4,4,4,4)) > area1 <- sample(factor(rep(c("a","b","c","d"),4))) > area2 <-

>>>>> "Jose" == Jose Ramon G Albert <toots at info.com.ph> writes: Jose> I have just downloaded this freeware version of R (which seems Jose> to be a clone of S). I was wondering if anyone knows where I Jose> could obtain R or S functions which provide nonparametric Jose> regression curves, e.g. kernel estimators or smoothing Jose> splines.

I now have versions of code that is destined (I believe) for 0.63 which is in a suitable state for comment. The files are at ftp://ftp.stats.ox.ac.uk/pub/R (Our www server is being moved, so may be intermittently down, but this ftp server should be stable.) All are R packages, for the moment for personal use only (no re-distribution). Use with 0.62.3 or 0.63 (although I am aware of some

Full_Name: Yves Rosseel Version: 2.2.1 OS: i686-pc-linux-gnu Submission from: (NULL) (157.193.116.152) Dear developers, For the single-model case, the anova.mlm() function does not seem to handle multi-parameter predictors (eg factors) correctly. A toy example illustrates the problem: Y <- cbind(rnorm(100),rnorm(100),rnorm(100)) A <- factor(rep(c(1,2,3,4), each=25)) fit <- lm(Y ~ A)

The following are now on CRAN: KernSmooth: version 2.2 of the code for Wand & Jones book on kernel smoothing. logspline: spline fits to log denisites, with automatic choice of smoothing. ppr: projection pursuit regression. rpart: recursive partitioning (CART-like) VR: Venables & Ripley libraries 5.3pl021 for 0.62.1 and in the devel section tree: a clone