13
$\begingroup$

What is the best place to ask questions about mathematical programming, between Math.SE and StackOverflow?

What I have in mind is the following type of questions:

  • Solving the word problem in a group given by presentation.
  • Plotting Cayley graphs of groups (e.g. a previous question I myself asked on SO a while ago).
  • Algorithms to plot curves/graphs of complicated/theoretical maps (e.g. visualizations of complex functions).
  • Given a theoretical algorithm from a paper (e.g. the Handle reduction algorithm to compare two braids), how to implement it?
  • Representations of higher-dimensional objects (e.g. plotting hypercubes in 3d-space).
  • Plotting implicit curves defined by algebraic equations in projective spaces over whatever field.
  • etc.

If the answer were to be either black or white, there would be two type of questions:

  1. The question is purely about the theoretical algorithm itself, and it belongs to Math.SE.
  2. The question is about a specific implementation and it belongs to SO.

However, there's a gray zone in between, where the question I linked above is a living example. It drew absolutely no attention, because people on SO don't know much about geometric group theory, and it is highly-specific knowledge. The question could even be qualified as useless for people to come, since people generically don't look for that type of things.

However, asking the question on Math.SE would've presented the same caveat: because it is purely about how to program the tree exploration algorithm in that particular scenario, it carries little-to-no interesting stuff from the mathematics point of vue.

Therefore, here I am asking the question: where should that gray zone belong to?

$\endgroup$
7
  • 1
    $\begingroup$ See also math.meta.stackexchange.com/questions/25495/… $\endgroup$
    – Asaf Karagila Mod
    Oct 2 at 16:59
  • 5
    $\begingroup$ Then there are also computational science, scicomp.SE, where numerical algorithms can be discussed, and computer science, cs.SE, where also questions of certain classes of finite structures (formal languages, graph theory, computational complexity and more) are discussed. $\endgroup$ Oct 4 at 15:54
  • $\begingroup$ @LutzLehmann I had no idea these existed, thanks for sharing! $\endgroup$ Oct 4 at 17:08
  • 1
    $\begingroup$ I also find the Operations Research site, or.SE, can be a good outlet for questions related to mathematical programming. $\endgroup$ Oct 5 at 1:22
  • $\begingroup$ I know your question is broader than a specific CAS, but there is also Mathematica SE $\endgroup$
    – Wojowu
    Oct 5 at 14:08
  • 5
    $\begingroup$ The word "mathematical programming" is often used to mean optimization (that is, mathematical optimization). That is an applied math topic. If you look at what the UCLA applied mathematicians are doing, for example, a bunch of them are doing optimization. I think applied math problems should always be welcome on math.stackexchange. Also, often it is applied mathematicians and not software engineers who know how to implement such algorithms. I would say that when in doubt, we should allow it. It's not as if the site is being overrun with mathematical programming questions. $\endgroup$
    – littleO
    Oct 11 at 4:49
  • $\begingroup$ Also, while we are on the topic, what about computer verification? For instance, is asking about proofs written in programming languages like Coq, Agda, Lean etc. more on topic for SO or Math.SE? $\endgroup$
    – Dair
    Oct 15 at 17:27

You must log in to answer this question.

Browse other questions tagged .