Interesting solution. Can someone explain how performant this is, i.e. how the short-circuiting here works, i.e. what happens when length of xs is smaller than n?

I already learned through this solution, that pattern matches are allowed to fail in some monads (like list and Maybe) :D

Just do what is fast in your particular language. Order of elements within a combination by definition does not matter.

Which is entirely correct.

The negative number issue is now fixed.

There's still a tab issue I forgot to fix, and it's now fixed.

It's not giving neg numbers. Closed

This issue no longer appears. Closed

same =(

Yeah, I've got exactly the same problem.
This may be fixed by replacing tabs with spaces on line 7 on tests file

I've got exactly the same problem

The following warning prevents me from submitting my otherwise correct solution:

/tmp/haskell116114-18-mo7fqm.mrxlpx2yb9/Main.hs:7:1: Warning:
Tab character

The test case doesn't consider the list has repeated element.
Like this, combinations 2 [1,1,1,1,1].

QuickCheck tests should not give negative numbers for group size (e.g. combinations -2 [1, 2, 3, 4])

Tests expect [[]] when inputting 0, but expect [] when inputting negative values or values greater than the length of the list.