Given an input string, output a random combination of any characters in the string, with an equal chance of each one occurring.
Example: given the input abcd (or any combination thereof of the four characters a,b,c,d) , there is an equal chance of outputting:
a b c d aa ab ac ad ba bb bc bd ca...
Must we remove (some) characters with a 50% chance and then remove a uniform number of characters in [1,Length(string)], or are we free to use other distributions (so long as we do not always do the same thing with the same input)?
I think you mean the code should have a (equal?) chance of outputting any subsequence but the word "can" could easily (although probably incorrectly) be interpreted as "may".
In future, we have a sandbox to help refine challenges before they're posted. Random challenges are notoriously hard to phrase properly, so don't feel too bad about this one! It's an interesting challenge, it just needs a bit of help to get it to the finish line.
Re my "(equal?)" - see my Jelly answer - it has a chance to produce any of the subsequences (with a 50% chance of the unchanged string), but the distribution of the number of characters removed is, itself, not uniform but rather the distribution across subsequences is - is that acceptable or not?
If the input is "aab", do you want equal probabilities of outputting "aa" and "ab"? There is only one was of getting "aa" (remove the "b"), but two ways of getting "ab" (remove the first "a", or remove the second "a").
There are a couple other questions in the Sandbox I could use some constructive feedback about: https://codegolf.meta.stackexchange.com/questions/2140/sandbox-for-proposed-challenges/17538#17538
Discussion on question by Flog Edoc: …
Discussion on question by Flog Edoc: …