The question Finding third side of a minimal triangle appears to be essentially a duplicate of https://stackoverflow.com/questions/22441855/triangle-circle-enclosing-a-set-of-points .

The stackoverflow question has detailed answers with references, for this nontrivial problem.

If it is a duplicate (which is not yet entirely clear, since the question is itself not entirely clear), should it be closed (I think yes), and if so, how? The option to close as duplicate does not work -- it says "The duplicate question must exist on Mathematics Stack Exchange".

(Sorry if this has been asked before, but I couldn't find it -- just close as a duplicate if it has!)

  • 2
    $\begingroup$ You can't. End of story. $\endgroup$
    – Asaf Karagila Mod
    Sep 3, 2016 at 12:00
  • 6
    $\begingroup$ Some related older discussions (I guess you can find more of them with some effort): What if we have duplicates “outside MSE”? and How should we handle cross-site duplicates? $\endgroup$ Sep 3, 2016 at 12:13
  • $\begingroup$ Not to mention the context makes the questions completely different. Have you noticed that both answers of the SO question are explaining that the algorithm is $O(n)$ and so on? Whereas this kind of consideration is non-existent in the math.SE thread. $\endgroup$ Sep 3, 2016 at 15:16
  • 2
    $\begingroup$ Whenever the software is unable to cope with a particular situation, or the default close reasons are insufficient, you can use the custom close reason, and just leave a comment as to why you are voting to close. You access this reason under Close -> Off Topic -> Other. I haven't looked at this particular situation, but that is the general method for voting to close. $\endgroup$ Sep 3, 2016 at 15:22
  • $\begingroup$ @CarlMummert: Thanks for the tip! $\endgroup$
    – Matt
    Sep 6, 2016 at 15:16
  • $\begingroup$ @Najib: I think the context is actually similar -- the OP is trying to write code to do it. When talking about an algorithm, mentioning its running time is a perfectly normal thing to do, even if the discussion is basically mathematical. (If you don't need an algorithm, then just say "Let T be the smallest triangle enclosing the points" and you are done!) $\endgroup$
    – Matt
    Sep 6, 2016 at 15:25
  • $\begingroup$ @Matt Have you not noticed that the time complexities of the algorithms aren't mentioned at all in the math.SE thread, whereas they are heavily emphasized in the SO thread? $\endgroup$ Sep 6, 2016 at 15:27
  • $\begingroup$ @Najib: Yes, but the "thread" you refer to is just me and one other person discussing the problem with the OP, and since neither of us are familiar with the O(n) algorithm, of course we did not mention it. It is surprising (to me) that there is an O(n) algorithm -- I would certainly have mentioned it too if I were talking about that algorithm. I feel that the SO and math.SE topic sets, although quite different in general, do have some overlap at their boundary. $\endgroup$
    – Matt
    Sep 6, 2016 at 15:53


Browse other questions tagged .