Hallo,
i tried writing a function to extract
all the odds ratio's from a ftable:
(+ p.adjust needs to build in)
So i tried the following:
ORCalcul <- function(m) {
or<-matrix(1:(length(m[,1])-1)*(length(m[1,])-1)*5,length(m[,1])-1,length(m[1,])-1)
for(i in 1:length(m[,1])-1) {
for(j in 1:length(m[1,])-1) {
or[i,j,1] <- attr(m,"row.vars")[[1]][i]
or[i,j,2] <- attr(m,"col.vars")[[1]][j]
or[i,j,3] <- attr(m,"row.vars")[[1]][i+1]
or[i,j,4] <- attr(m,"col.vars")[[1]][j+1]
or[i,j,5] <-(m[i,j]*m[i+1,j1])/(m[i1,j]*m[i,j1])
}
}
ORCalcul <- c(or)
}
## which doenst work, because of the way R works with matrices,
## i am justed to other syntax, using the indices of a matrix
## so first question: i don't think there is way in R to work with
matrix m = new matrix(10,5,45)
# m being a three dim matrix eg?
# so i tried, and tried, finally
ORCalcul <- function(m) {
t
for(i in 1:length(m[,1])-1) {
for(j in 1:length(m[1,])-1) {
row1 <- attr(m,"row.vars")[[1]][i]
col1 <- attr(m,"col.vars")[[1]][j]
row2 <- attr(m,"row.vars")[[1]][i+1]
col2 <- attr(m,"col.vars")[[1]][j+1]
or <- (m[i,j]*m[i+1,j+1])/(m[i+1,j]*m[i,j+1])
ORCalculij <- c(row1, col1, row2, col2, or)
append(t,ORCalculij)
}
}
ORCalcul <- t
ORCalcul
}
# which unfornately only gives as output:, so doesnt work either
[,1]
[1,] NA
# the only variant i wrote that gave some output,
# but logically, only the last OR, is:
ORCalcul <- function(m) {
for(i in 1:length(m[,1])-1) {
for(j in 1:length(m[1,])-1) {
row1 <- attr(m,"row.vars")[[1]][i]
col1 <- attr(m,"col.vars")[[1]][j]
row2 <- attr(m,"row.vars")[[1]][i+1]
col2 <- attr(m,"col.vars")[[1]][j+1]
or <- (m[i,j]*m[i+1,j+1])/(m[i+1,j]*m[i,j+1])
ORCalcul <- c(row1, col1, row2, col2, or)
# putting an append here, just gives me the code
# back..
}
}
ORCalcul
}
# what should i do? Suggestions?
(for the algoritmic logic, i need two more nested loops, that is not the
problem)
mvg,
Wim
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
