similar to: LiblineaR: read/write model files?

Thibault, It would be nice if LiblineaR() accepted data in the form of a sparse matrix (it does not accept whatever e1071::read.matrix.csr returns). It would also be nice if there were functions to read/write files in the native liblinear file format; I am sure the original liblinear library provides at least the input code. Thanks! -- Sam Steingold (http://sds.podval.org/) on Ubuntu 12.04

I've been trying to get some linear classifiers (LiblineaR, kernlab, e1071) to work with a sparse matrix of feature data. In the case of LiblineaR and kernlab, it seems I have to coerce my data into a dense matrix in order to train a model. I've done a number of searches, read through the manuals and vignettes, but I can't seem to see how to use either of these packages with sparse

suppose I have two sets of vectors: x1,x2,...,xN and y1,y2,...,yN. I want N correlations: cor(x1,y1), cor(x2,y2), ..., cor(xN,yN). my sets of vectors are arranged as data frames x & y (vector=column): x <- data.frame(a=rnorm(10),b=rnorm(10),c=rnorm(10)) y <- data.frame(d=rnorm(10),e=rnorm(10),f=rnorm(10)) cor(x,y) returns a _matrix_ of all pairwise correlations: cor(x,y)

Hello List, I am having a great trouble using svm function in e1071 package. I have 4gb of data that i want to use to train svm. I am using Amazon cloud, my Amazon Machine Image(AMI) has 34.2 GB of memory. my R process was killed several times when i tried to use 4GB of data for svm. Now I am using a subset of that data and it is only 1.4 GB. i remove all unnecessary objects before calling

Hi, to count vector elements with some property, the standard idiom seems to be length(which): --8<---------------cut here---------------start------------->8--- x <- c(1,1,0,0,0) count.0 <- length(which(x == 0)) --8<---------------cut here---------------end--------------->8--- however, this approach allocates and discards 2 vectors: a logical vector of length=length(x) and an

How do I convert a list to a matrix? --8<---------------cut here---------------start------------->8--- list(c(50000, 101), c(1e+05, 46), c(150000, 31), c(2e+05, 17), c(250000, 19), c(3e+05, 11), c(350000, 12), c(4e+05, 25), c(450000, 19), c(5e+05, 16)) as.matrix(a) [,1] [1,] Numeric,2 [2,] Numeric,2 [3,] Numeric,2 [4,] Numeric,2 [5,] Numeric,2 [6,] Numeric,2 [7,]

At the end of a for loop its variables are still present: for (i in 1:10) { x <- vector(length=100000000) } ls() will print "i" and "x". this means that at the end of the for loop body I have to write rm(x) gc() is there a more elegant way to handle this? Thanks. -- Sam Steingold (http://sds.podval.org/) on Ubuntu 12.04 (precise) X 11.0.11103000

The following code: --8<---------------cut here---------------start------------->8--- > myfun <- function (x) list(x=x,y=x*x) > z <- as.data.frame(do.call(rbind,lapply(1:3,function(x) c(a=paste("a",x,sep=""),as.list(unlist(list(b=myfun(x),c=myfun(x*x*x)))))))) > z a b.x b.y c.x c.y 1 a1 1 1 1 1 2 a2 2 4 8 64 3 a3 3 9 27 729

I have a large data.frame Z (2,424,185,944 bytes, 10,256,441 rows, 17 columns). I want to get the result of table(aggregate(Z$V1, FUN = length, by = list(id=Z$V2))$x) alas, aggregate has been running for ~30 minute, RSS is 14G, VIRT is 24.3G, and no end in sight. both V1 and V2 are characters (not factors). Is there anything I could do to speed this up? Thanks. -- Sam Steingold

I have the following code: --8<---------------cut here---------------start------------->8--- d <- rep(10,10) for (i in 1:100) { a <- sample.int(length(d), size = 2) if (d[a[1]] >= 1) { d[a[1]] <- d[a[1]] - 1 d[a[2]] <- d[a[2]] + 1 } } --8<---------------cut here---------------end--------------->8--- it does what I want, i.e., modified vector d 100 times.

I find myself doing --8<---------------cut here---------------start------------->8--- tab <- table(...) tab <- tab[tab > 0] tab <- sort(tab,decreasing=TRUE) --8<---------------cut here---------------end--------------->8--- all the time. I am wondering if the "drop 0" (and maybe even sort?) can be effected by some magic argument to table() which I fail to discover

why? > rle Run Length Encoding lengths: int [1:1650061] 2 2 8 2 4 5 6 3 26 46 ... values : chr [1:1650061] "4bbf9e94cbceb70c BG bg" "4fbbf2c67e0fb867 SK sk" ... > as.data.frame(rle) Error in as.data.frame.default(vertices.rle) : cannot coerce class '"rle"' into a data.frame it seems that rle.df <-

I need an analogue of "uniq -c" for a data frame. xtabs(), although dog slow, would have footed the bill nicely: --8<---------------cut here---------------start------------->8--- > x <- data.frame(a=1:32,b=1:32,c=1:32,d=1:32,e=1:32) > system.time(subset(as.data.frame(xtabs( ~. , x )), Freq != 0 )) user system elapsed 12.788 4.288 17.224 --8<---------------cut

What do people use for SNA/graph analysis in R? So far I have been using igraph (it implements the Louvain community detection algorithm as multilevel.community, which is the killer feature for me). However, igraph is severely lacking in visualization, which I also need. graphviz & gephi are alleged to be good at visualization, but, apparently, not so for analysis (specifically, community

I have a long numeric vector v (length N) and I want create a shorter vector of length N/k consisting of sums of k-subsequences of v: v <- c(1,2,3,4,5,6,7,8,9,10) N=10, k=3 ===> [6,15,24,10] I can, of course, iterate: > w <- vector(mode="numeric",length=ceiling(N/k)) > for (i in 1:length(w)) w[i] <- sum(v(i*k:(i+1)*k)) (modulo boundary conditions) but I wonder if

I have a graph with edge and vertex weights, stored in two data frames: --8<---------------cut here---------------start------------->8--- vertices <- data.frame(vertex=c("a","b","c","d"),weight=c(1,2,1,3)) edges <-

summary() for a factor prints: ColName SNDK : 72 VXX : 36 MWW : 30 ACI : 28 FRO : 28 (Other):1801 it would have been much more useful if it additionally printed frequency stats as if by summary(aggregate(frame$ColName,by=list(frame$ColName),FUN=length)$x) -- Sam Steingold (http://sds.podval.org/) on CentOS release 5.3 (Final) http://jihadwatch.org

How do I concatenate two vectors of factors? --8<---------------cut here---------------start------------->8--- > a <- factor(5:1,levels=1:9) > b <- factor(9:1,levels=1:9) > str(c(a,b)) int [1:14] 5 4 3 2 1 9 8 7 6 5 ... > str(unlist(list(a,b),use.names=FALSE)) Factor w/ 9 levels "1","2","3","4",..: 5 4 3 2 1 9 8 7 6 5 ...

Hi, what does this mean? --8<---------------cut here---------------start------------->8--- > graph <- graph.data.frame(merged[!v,], vertices=ve, directed=FALSE) cedta decided 'igraph' wasn't data.table aware cedta decided 'igraph' wasn't data.table aware cedta decided 'igraph' wasn't data.table aware cedta decided 'igraph' wasn't

Using Rgraphviz, I draw the undirected graph with vertices A,B,C and D and edges A:B, B:C, C:D, D:A, A:C. I want the vertices A and B to be red and C and D to be blue. The problem is the following: I want the edges A:B and B:C to be green and the edges C:D and C:A to be yellow, while the edge A:C can have the default colour black. I assume that I have to specify this using the edgeAttrs-argument