See if this helps: http://finzi.psych.upenn.edu/R/Rhelp02a/archive/33330.html Have no idea if it works w/o change in S-PLUS though. Andy -- Notice: This e-mail message, together with any attachments,
I guess the simplest approach is to read the data in pieces, and accumulate the SSCP matrix. Then use the Cholesky decomposition to get the estimates. You're quite literally doing the regression `by
I'll chip in my $0.02: 1. I find Insightful's marketing blurb for the big data library in the Enterprise Developer version a bit puzzling: I just check the price of a dual Opteron 275 server with 8GB
Assuming the data is in a data frame, and the `year' and `idcode' are of type `character', the following works in R: 8025 0987 1980 0.9880 0.9880 1981 0.9877 0.9877 1982 0.8766 0.8766 Andy -- Notice:
Class labels are not the only thing produced by the tree: It also outputs estimated class probabilities. Those `0' terminal nodes probably all have different estimated probabilities. Andy -- Notice:
Perhaps slightly simpler: library(Design) rcs <- rcs rcspline.eval <- rcspline.eval detach() Andy -- Notice: This e-mail message, together with any attachments, contains information of Merck & Co.,
One trick is to use `index' as the row indices, and generate column indices, then use both: mytest2 <- function() { m <- 3 n <- 4 rindex <- c(1,1,1,2,2,2,2,3,3) cindex <- unlist(lapply(table(rindex),
If your columns are already sorted in the right way, just make the matrix into an array and aperm() as needed. Here's an example: A1.B1.C1 A2.B1.C1 A1.B2.C1 A2.B2.C1 A1.B1.C2 A2.B1.C2 A1.B2.C2 A2.B2.
You need to make `a' and `b' factors. Andy -- Notice: This e-mail message, together with any attachments, contains information of Merck & Co., Inc. (One Merck Drive, Whitehouse Station, New Jersey, U
you elaborate on what part(s) of the output that are so different beyond your reasonable explanation? You fitted two conceptually very different models. Why would you expect the result to be similar
I suppose by nnet() you meant the function in the MASS library. If so, that's support software for the book `MASS', which contains discussion and examples. Have you tried that? (Google is a very usef
Is this the link you tried? http://www.stats.ox.ac.uk/pub/MASS3/Exegeses.pdf Andy Hello S+ users, Also, i was viewing the help file for 'multicomp' to perform some post-hoc comparisons for a nlme ana