I use LETTERS most of the time for my factors but today I tried to go beyond 26 characters:
LETTERS[1:32]
Expecting there to be an automati
Here's another addition to the list. This seems a bit faster than Gregor's (comparison done on my computer - using length.out = 1e6
his took 12.88 seconds, mine was 6.2), and can also be extended indefinitely. The flip side is that it's 2 functions, not just 1.
make.chars <- function(length.out, case, n.char = NULL) {
if(is.null(n.char))
n.char <- ceiling(log(length.out, 26))
m <- sapply(n.char:1, function(x) {
rep(rep(1:26, each = 26^(x-1)) , length.out = length.out)
})
m.char <- switch(case,
'lower' = letters[m],
'upper' = LETTERS[m]
)
m.char <- LETTERS[m]
dim(m.char) <- dim(m)
apply(m.char, 1, function(x) paste(x, collapse = ""))
}
get.letters <- function(length.out, case = 'upper'){
max.char <- ceiling(log(length.out, 26))
grp <- rep(1:max.char, 26^(1:max.char))[1:length.out]
unlist(lapply(unique(grp), function(n) make.chars(length(grp[grp == n]), case = case, n.char = n)))
}
##
make.chars(5, "lower", 2)
#> [1] "AA" "AB" "AC" "AD" "AE"
make.chars(5, "lower")
#> [1] "A" "B" "C" "D" "E"
make.chars(5, "upper", 4)
#> [1] "AAAA" "AAAB" "AAAC" "AAAD" "AAAE"
tmp <- get.letters(800)
head(tmp)
#> [1] "A" "B" "C" "D" "E" "F"
tail(tmp)
#> [1] "ADO" "ADP" "ADQ" "ADR" "ADS" "ADT"
Created on 2019-03-22 by the reprex package (v0.2.1)
You can make what you want like this:
LETTERS2<-c(LETTERS[1:26], paste0("A",LETTERS[1:26]))
A little late to the party, but I want to play too.
You can also use sub
, and sprintf
in place of paste0
and get a length 702 vector.
c(LETTERS, sapply(LETTERS, sub, pattern = " ", x = sprintf("%2s", LETTERS)))