On 4/16/2009 2:58 PM, Juergen Rose wrote:
The second beginner question. I want to create a new dataframe, where
each element of the original dataframe is translated to 1 if it was "+",
to 0 if it was "-" to -1 otherwise. I could do with:
Lines <- "a b c d
+ - + +
+ + + -
+ 1 - '+ '
- + + +
+ N - +"
DF <- read.table(textConnection(Lines), header = TRUE)
cnames <- colnames(DF)
nrow <-length(rownames(DF))
nc <- length(cnames)
NDF <- data.frame(matrix(c(rep(0,nc*nrow)),ncol=nc))
for (i in 1:length(cnames)) {
NDF[,i] <- sapply(DF[,i],function(x) if (x=="+") {1} else {if (x=="-")
{0} else {-1}} )
}
colnames(NDF) <- cnames
But this is shure one loop too much. Please give me the R way solution.
library(car)
DF[] <- lapply(DF, function(x){recode(x, "'+'=1; '-'=0; else=-1")})