similar to: probabilistic neural networks

Hi, I was trying to use the nnet library and am not sure of whats going on. I am calling the nnet function as: n <- nnet(x,y,size=3,subset=sets[[1]], maxit=200) Where x is a 272x4 matrix of observations (examples) and y is a 272x1 matrix of target values. However when I look at nnet$residuals they are off by two orders of magnitude (compared to the output from neural network code that I

Hi, this question is probably very obvious but I just cant see where I might be going wrong. I'm using the lm() function to generate a linear model and then make predictions using a different set of data. To generate the model I do (tdata & pdata are matrices of observations and parameters, tdepv, pdepv are response vectors) x <- as.data.frame(tdata) x$tdepv <- tdepv

Hi, we have some code that does variable selection with a genetic algorithm or simulated annealing, using a linear regression routine or neural network as the objective function. This code is a mixture of fortran and C. The code is more than 15 years old and I am planning to rework it. Though a C rewrite would be good for efficiency, I would like to protoype it quickly in R and see how it

Hi, I'm trying to take a set of rows and columns out of a matrix. I hve been using the index aray approach. My overll matrix is X and is 179 x 65. I want to take out 4 columns and 161 rows. Thus I made a 161 x 2 array I and filled it up with the row,col indices. However doing, X[ I ] gives me a vector of the extracted elements. Is there anyway I can coerce this into a 161 x 4 matrix?

Hi, I've been generating linear models with lm(). I wanted to look at the VIF's for the coefficient. Using the vif() function from the package Design, I would get unusually high VIF's. However using vif() from the car package I get more reasonable values (ie in line with the quality of the model). What is the difference between the two vif functions? (I dont have access to the

Hi, is there a function in R that will give me the variances of a predicted values obtained using predict.lm(). If no function is available I would need to calculate them myself - which involves taking the inverse of X'X (' indicating transpose) where X is my model matrix. I know that calculating an inverse directly is not a good idea in general - could anybody suggest a way around

Hi, I was wondering if there are any packages available that can represent mathematical graphs along with functions to manipulate them? Google did'nt turn up anything ( I may be asking too much of R :-/ ) Thanks, ------------------------------------------------------------------- Rajarshi Guha <rxg218 at psu.edu> <http://jijo.cjb.net> GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04

Hi, I have an integer matrix consisting of 1's and 0's and I would like to convert this to a data.frame where each column of the matrix becomes a factor variable. Now, some columns of the matrix have only 1's or only 0's as a result there is only 1 level for those columns in the data.frame. However it is required that each factor have 2 levels. So my solution is: m <-

Hi, I'm trying to translate some Matlab code to R and I'm trying to implement the behavior of Matlab's sort() which when applied to a matrix will sort the columns returning the column sorted matrix as well as a matrix of permuted indices. Doing: > x <- matrix(c(3,4,2,6,3,4,8,7,5), nr=3) > x [,1] [,2] [,3] [1,] 3 6 8 [2,] 4 3 7 [3,] 2 4 5

Hi, I have an apply statement that looks like: > check.cols <- function(v1, v2) { + return( identical(v1,v2) ); + } > x [,1] [,2] [,3] [1,] 1 3 3 [2,] 4 5 4 [3,] 2 7 6 > apply(x, c(2), check.cols, v2=c(7,8,9)) [1] FALSE FALSE FALSE Is it possible to make the function check.cols() inline to the apply statement. Some thing like this: apply(x,

Hi, I'm making a plot in which the title takes up two lines. The title contains a subscript but when I look at the plot it does'nt seem tocome out properly. The code I'm using is: xtxt = expression(paste('Observed -log( ', IC[50], ' )')) ytxt = expression(paste('Predicted -log( ', IC[50], ' )')) mtxt = expression(paste('Plot of Observed

Hi, I have an old version of the Rggobi package which now wont load (R 2.0.0). Looking for the package on CRAN does'nt turn up anything and the links on http://www.ggobi.org/RSggobi.html don't work. Does anybody know where I can get the latest version of RSggobi or Rggobi? Thanks, ------------------------------------------------------------------- Rajarshi Guha <rxg218 at

Hi, I am working on a Java project that will run primarily on Windows. As part of the project I would like to use the R engine. However, from what I have seen on this list & the SJava list, installing SJava on Windows can be troublesome. As a result I was wondering if anybody had used DCOM to access R from a Java program on Windows (if it is all possible). Any pointers to documents and/or

Hi, I have a rectangular matrix and I need to check whether any columns are identical or not. Currently I'm looping over the columns and checking each column with all the others with identical(). However, as experience has shown me, getting rid of loops is a good idea :) Would anybody have any suggestions as to how I could do this job more efficiently. (It would be nice to know which

Hi, I'm using nnet to work on a 2 class classification problem. The result of my code is data.frame of true class, predicted class and associated probability. One way of summarizing the data is by a confusion matrix. However are there any graphical ways I could represent the data - specifically, I'd like to show the probabilities associated with each member of my prediction set? (I

Hi, I had some Java code that worked with SJava 0.65 under R 1.8.*. I'm trying to get it to work with R 2.0.0. My JVM is Sun 1.5.0 (running on Fedora Core 2) I downloaded and installed SJava 0.65 in my personal directory using R CMD INSTALL SJava-0.65.tar.gz -l ~/src/Rlibrary and before starting R I sourced RJava.bsh. However after loading the SJava library if do javaConfig() I get

Hi, is this feature available in R? I plot some data using plot(). Clicking on a point I'd like information regarding that point. I realize that I could use locator() but this gives me the X & Y values. Is it possible to somehow use locator to give more information regarding the point chosen - say, which vector the point was taken from. It seems that R itself does not do this - is

Hi, I have some code that looks like: dftc <- df[sets$tcset,] pt <- numeric(nrow(dftc)) sub <- 1:nrow(dftc) for (i in 1:nrow(dftc)) { n <- nnet( fmla, data=dftc, weights=wts, subset=sub[-i], size=4, decay=0.01) pt[i] <- predict( n, dftc[ i, ], type='class' ) } However running this give me the error: Error in eval(expr, envir, enclos) :

Hi, there is a quote in the fortunes package: To paraphrase provocatively, `machine learning is statistics minus any checking of models and assumptions'. -- Brian D. Ripley (about the difference between machine learning and statistics) useR! 2004, Vienna (May 2004) Was this statement made in a presentation? If so, would it be possible to get a reference for the presentation?

Hi, I have a number of web services that run on a server and can be accessed via SOAP as well as by creating a specific URL. Now, for certain services, the argument that I need to send is about 10 to 20 characters and the URL approach works fine. However if the argument is a few kilobytes I would rather create a SOAP packet and send that off. So my first question is: is there a package that