Haskell：从x元素列表中创建唯一的n元组

Question

I have a list of x elements from which I would like to have all possible unique n-tuples.

So I basically am looking for an implementation of

nub . map (take n) . permutations

that doesn't unnecessarily create the duplicates.

For me, this looks like a function that has a good chance of being defined already somewhere.

Is this the case?

Answer 1

Do you mean something like this?

import Data.List (permutations)

choose n list = concatMap permutations $ choose' list [] where
  choose' []     r = if length r == n then [r] else []
  choose' (x:xs) r | length r == n = [r]
                   | otherwise     = choose' xs (x:r) 
                                  ++ choose' xs r

Output:

*Main> choose 2 [0..5]
[[1,0],[0,1],[2,0],[0,2],[3,0],[0,3],[4,0],[0,4],[5,0],[0,5],[2,1]
,[1,2],[3,1],[1,3],[4,1],[1,4],[5,1],[1,5],[3,2],[2,3],[4,2],[2,4]
,[5,2],[2,5],[4,3],[3,4],[5,3],[3,5],[5,4],[4,5]]

Answer 2

Hayoo is very helpful for finding functions. With this query , I found the permutation package, which seems it might do what you want, although I haven't examined the implementation in detail.

But it might be simpler to use one of the solutions discussed here on Stack Overflow. A search turned up some relevant posts. You might look at this discussion of the implementation of the permutations function in Data.List, and modify it to meet your needs.

Answer 3

Since you've asked for tuples, here is a solution exploiting the fact that list is an applicative functor:

Prelude> let list = [0..4]
Prelude> import Control.Applicative
Prelude Control.Applicative> (,) <$> list <*> list
[(0,0),(0,1),(0,2),(0,3),(0,4),(1,0),(1,1),(1,2),...
Prelude Control.Applicative> (,,) <$> list <*> list <*> list
[(0,0,0),(0,0,1),(0,0,2),(0,0,3),(0,0,4),(0,1,0),(0,1,1),...

However, since tuples of different arities are different types by definition, it's impossible to write a universal function which will produce results of different arities based on some n parameter.

Although I must note that this could be solved using some advanced type-level programming techniques by introducing a type-class and instances for type-level naturals representing the arities of tuples you wanted to support, but I'm sure this would be an overkill. Besides, utilizing type-naturals would also be redundant, since all the required information could be determined from the result type, so a simple type class and instances for all tuples you wanted to support would be enough.

Answer 4

You can use Math.Combinatorics.Multiset ( cabal install multiset-comb ), which will avoid duplicates in the output list even if there are duplicates in the input list:

import Math.Combinatorics.Multiset  (fromList, kSubsets, permutations)

permute :: Ord a => Int -> [a] -> [[a]]
permute n = concatMap permutations . kSubsets n . fromList

main = print $ permute 2 [1, 1, 2, 3]

Produces:

[[2,3],[3,2],[1,3],[3,1],[1,2],[2,1],[1,1]]