IP-Address

Allan Engelhardt wrote:
  

IP addresses are very (very!) difficult to parse and sort correctly
because there are all sorts of supported formats.  Try to use something
like PostgreSQL instead: it is already implemented there.  But if you
are sure all your data is of the n.n.n.n form, then something along the
lines of the following should basically work (I have chosen some more
interesting IP addresses for this):


a <- data.frame(cbind(id=c(138,138,138,138),
                     rank=c(29746,29746,29746,29746),
                     color=c("yellow","red","blue","red"),
                     status=c("no","yes","yes","no"),
                    
ip=c("162.131.58.26","2.131.58.16","2.2.58.10","162.131.58.17")))
a
#    id  rank  color status            ip
# 1 138 29746 yellow     no 162.131.58.26
# 2 138 29746    red    yes   2.131.58.16
# 3 138 29746   blue    yes     2.2.58.10
# 4 138 29746    red     no 162.131.58.17
x <- matrix(unlist(lapply(strsplit(as.character(a$ip), ".", fixed=TRUE),
as.integer)),
           ncol=4, byrow=TRUE)
a[order(x[,1],x[,2],x[,3],x[,4]),]
#    id  rank  color status            ip
# 3 138 29746   blue    yes     2.2.58.10
# 2 138 29746    red    yes   2.131.58.16
# 4 138 29746    red     no 162.131.58.17
# 1 138 29746 yellow     no 162.131.58.26


Getting rid of the conversions including the matrix(unlist) combo is
left as an exercise (it's too hot here....)
    

Here's one way:

con <- textConnection(as.character(a$ip))
o <- do.call(order,read.table(con,sep="."))
close(con)
a[o,]