R - shuffle a list preserving element sizes - r

In R, I need an efficient solution to shuffle the elements contained within a list, preserving the total number of elements, and the local element sizes (in this case, each element of the list is a vector)
a<-LETTERS[1:6]
b<-LETTERS[6:10]
c<-LETTERS[c(9:15)]
l=list(a,b,c)
> l
[[1]]
[1] "A" "B" "C" "D" "E" "F"
[[2]]
[1] "F" "G" "H" "I" "J"
[[3]]
[1] "I" "J" "K" "L" "M" "N" "O"
The shuffling should randomly select the letters of the list (without replacement) and put them in a random position of any vector within the list.
I hope I have been clear! Thanks :-)

you may try recreating a second list with the skeleton of the first, and fill it with all the elements of the first list, like this:
u<-unlist(l)
l2<-relist(u[sample(length(u))],skeleton=l)
> l2
[[1]]
[1] "F" "A" "O" "I" "S" "Q"
[[2]]
[1] "R" "P" "K" "F" "G"
[[3]]
[1] "A" "N" "M" "J" "H" "G" "E" "B" "T" "C" "D" "L"
Hope this helps!

Like this...?
> set.seed(1)
> lapply(l, sample)
[[1]]
[1] "B" "F" "C" "D" "A" "E"
[[2]]
[1] "J" "H" "G" "F" "I"
[[3]]
[1] "J" "M" "O" "L" "N" "K" "I"

Related

Print vowels from the vector

I need to execute the vowels from the LETTERS R build-in vector
"A", "E", etc.
> LETTERS
[1] "A" "B" "C" "D" "E" "F" "G" "H" "I" "J" "K" "L" "M" "N" "O" "P" "Q" "R" "S" "T" "U"
"V" "W" "X"
[25] "Y" "Z"
Maybe, someone knows how to do it with if() or other functions. Thank you in advance.
Looks like you need extract vowels, does this work:
> vowels <- c('A','E','I','O','U')
> LETTERS[sapply(vowels, function(ch) grep(ch, LETTERS))]
[1] "A" "E" "I" "O" "U"
>

R: sample function repeats same results

I ran into something very weird with sample(). If I run the following line 5 times at the start of a session (in either RStudio or R), I would get the following results.
sample(letters,5,replace=TRUE)
[1] "b" "y" "d" "p" "n"
[1] "v" "n" "i" "s" "s"
[1] "d" "q" "a" "m" "x"
[1] "w" "s" "u" "h" "e"
[1] "b" "y" "g" "s" "e"
But if I restart the console and run it 5 times at the beginning of a new session, I would somehow get the same results -- every time. Is sample() (which I believe uses Mersenne Twister by default) supposed to do this? What should I do instead to get results that don't actually repeat?
set.seed(123)
> sample(letters,5,replace=TRUE)
[1] "h" "u" "k" "w" "y"
> sample(letters,5,replace=TRUE)
[1] "b" "n" "x" "o" "l"
> sample(letters,5,replace=TRUE)
[1] "y" "l" "r" "o" "c"
> sample(letters,5,replace=TRUE)
[1] "x" "g" "b" "i" "y"
> sample(letters,5,replace=TRUE)
[1] "x" "s" "q" "z" "r"
If you start a new session and change the set.seed value, you will get different results.
> set.seed(456)
> sample(letters,5,replace=TRUE)
[1] "c" "f" "t" "w" "u"
> sample(letters,5,replace=TRUE)
[1] "i" "c" "h" "g" "k"
> sample(letters,5,replace=TRUE)
[1] "j" "f" "t" "v" "p"
> sample(letters,5,replace=TRUE)
[1] "q" "v" "l" "s" "h"
> sample(letters,5,replace=TRUE)
[1] "e" "s" "x" "l" "v"
Hope that helps.

R: Efficient way for spreading vectors

Is there an efficient way of programming to solve the following task?
Imagine the following vector:
A<-[a,b,c...k]
And would like to spread it the following way:
Let‘s start with e.g. n=2
B<-[a,a,b,b,c...,k,k]
And now n=4 or any number greater 1
C<-[a,a,a,a,b,...,k,k,k,k]
To solve it via loops seems kind of easy, but is there any function or vector based operation I missed/could use? A tidyverse solutions (for using it in a pipe) would be the best solution for me.
(It is hard to do research on this task as I am a newbie in R and don‘t the correct terms to search for. Any help would be helpful.)
Let
A <- letters[1:11]
A
[1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k"
If you use function rep with argument each, you get what you want:
rep(A, each=2)
[1] "a" "a" "b" "b" "c" "c" "d" "d" "e" "e" "f" "f" "g" "g" "h" "h" "i" "i" "j"
[20] "j" "k" "k"
rep(A, each=3)
[1] "a" "a" "a" "b" "b" "b" "c" "c" "c" "d" "d" "d" "e" "e" "e" "f" "f" "f" "g"
[20] "g" "g" "h" "h" "h" "i" "i" "i" "j" "j" "j" "k" "k" "k"
An option is to use rep with argument times = 2 or 4 and then sort the result. Another option is to use mapply and then c operator.
c(mapply(rep, 2 ,A)) # OR sort(rep(A, times = 2))
#[1] "a" "a" "b" "b" "c" "c" "d" "d" "e" "e" "f" "f" "g" "g" "h" "h" "i" "i" "j" "j"
#[21] "k" "k"
c(mapply(rep,A, 4)) #OR sort(rep(A, times = 2))
#[1] "a" "a" "a" "a" "b" "b" "b" "b" "c" "c" "c" "c" "d" "d" "d" "d" "e" "e" "e" "e"
#[21] "f" "f" "f" "f" "g" "g" "g" "g" "h" "h" "h" "h" "i" "i" "i" "i" "j" "j" "j" "j"
#[41] "k" "k" "k" "k"

Matching between a vector and multiple vectors in a list in R

I have a list of vectors such as:
>list
[[1]]
[1] "a" "m" "l" "s" "t" "o"
[[2]]
[1] "a" "y" "o" "t" "e"
[[3]]
[1] "n" "a" "s" "i" "d"
I want to find the matches between each of them and the remaining (i.e. between the 1st and the other 2, the 2nd and the other 2, and so on) and keep the couple with the highest number of matches. I could do it with a "for" loop and intersect by couples. For example
for (i in 2:3) { intersect(list[[1]],list[[i]]) }
and then save the output into a vector or some other structure. However, this seems so inefficient to me (given than rather than 3 I have thousands) and I am wondering if R has some built-in function to do that in a clever way.
So the question would be:
Is there a way to look for matches of one vector to a list of vectors without the explicit use of a "for" loop?
I don't believe there is a built-in function for this. The best you could try is something like:
lsts <- lapply(1:5, function(x) sample(letters, 10)) # make some data (see below)
maxcomb <- which.max(apply(combs <- combn(length(lsts), 2), 2,
function(ix) length(intersect(lsts[[ix[1]]], lsts[[ix[2]]]))))
lsts <- lsts[combs[, maxcomb]]
# [[1]]
# [1] "m" "v" "x" "d" "a" "g" "r" "b" "s" "t"
# [[2]]
# [1] "w" "v" "t" "i" "d" "p" "l" "e" "s" "x"
A dump of the original:
[[1]]
[1] "z" "r" "j" "h" "e" "m" "w" "u" "q" "f"
[[2]]
[1] "m" "v" "x" "d" "a" "g" "r" "b" "s" "t"
[[3]]
[1] "w" "v" "t" "i" "d" "p" "l" "e" "s" "x"
[[4]]
[1] "c" "o" "t" "j" "d" "g" "u" "k" "w" "h"
[[5]]
[1] "f" "g" "q" "y" "d" "e" "n" "s" "w" "i"
datal <- list (a=c(2,2,1,2),
b=c(2,2,2,4,3),
c=c(1,2,3,4))
# all possible combinations
combs <- combn(length(datal), 2)
# split into list
combs <- split(combs, rep(1:ncol(combs), each = nrow(combs)))
# calculate length of intersection for every combination
intersections_length <- sapply(combs, function(y) {
length(intersect(datal[[y[1]]],datal[[y[2]]]))
}
)
# What lists have biggest intersection
combs[which(intersections_length == max(intersections_length))]

Letter "y" comes after "i" when sorting alphabetically

When using function sort(x), where x is a character, the letter "y" jumps into the middle, right after letter "i":
> letters
[1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s" "t"
[21] "u" "v" "w" "x" "y" "z"
> sort(letters)
[1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "y" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s"
[21] "t" "u" "v" "w" "x" "z"
The reason may be that I am located in Lithuania, and this is "lithuanian-like" sorting of letters, but I need normal sorting. How do I change the sorting method back to normal inside R code?
I'm using R 2.15.2 on Win7.
You need to change the locale that R is running in. Either do that for your entire Windows install (which seems suboptimal) or within the R sessions via:
Sys.setlocale("LC_COLLATE", "C")
You can use any other valid locale string in place of "C" there, but that should get you back to the sort order for letters you want.
Read ?locales for more.
I suppose it is worth noting the sister function Sys.getlocale(), which queries the current setting of a locale parameter. Hence you could do
(locCol <- Sys.getlocale("LC_COLLATE"))
Sys.setlocale("LC_COLLATE", "lt_LT")
sort(letters)
Sys.setlocale("LC_COLLATE", locCol)
sort(letters)
Sys.getlocale("LC_COLLATE")
## giving:
> (locCol <- Sys.getlocale("LC_COLLATE"))
[1] "en_GB.UTF-8"
> Sys.setlocale("LC_COLLATE", "lt_LT")
[1] "lt_LT"
> sort(letters)
[1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "y" "j" "k" "l" "m" "n"
[16] "o" "p" "q" "r" "s" "t" "u" "v" "w" "x" "z"
> Sys.setlocale("LC_COLLATE", locCol)
[1] "en_GB.UTF-8"
> sort(letters)
[1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o"
[16] "p" "q" "r" "s" "t" "u" "v" "w" "x" "y" "z"
> Sys.getlocale("LC_COLLATE")
[1] "en_GB.UTF-8"
which of course is what #Hadley's Answer shows with_collate() doing somewhat more succinctly once you have devtools installed.
If you want to do this temporarily, devtools provides the with_collate function:
library(devtools)
with_collate("C", sort(letters))
# [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s"
# [20] "t" "u" "v" "w" "x" "y" "z"
with_collate("lt_LT", sort(letters))
# [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "y" "j" "k" "l" "m" "n" "o" "p" "q" "r"
# [20] "s" "t" "u" "v" "w" "x" "z"

Resources