Given $P = \{x \in\mathbb R^n \mid a_1x_1 + \ldots + a_nx_n = \text{constant}\}$, $(a_1, \ldots , a_n) \ne 0$. Can $P$ be a polytope? I think that with $N = 1$, $P$ is a point. Can a point in $\mathbb R^1$ be a polytope?

Thank you all!

  • $\begingroup$ $P$ looks more like a hyperplane. $\endgroup$ – Earthliŋ Jun 17 '13 at 15:41
  • $\begingroup$ What if N = 1 ? $\endgroup$ – Fabio Carello Jun 17 '13 at 15:57
  • $\begingroup$ Then you have a 0-dimensional hyperplane, i.e. a point, which you can interpret as a 0-dimensioal polytope if you like. Your formula for $P$ only gives you polytopes when $n=1$, though. For $n>1$, the formula defines a hyperplane, which won't be a polytope. $\endgroup$ – Earthliŋ Jun 17 '13 at 16:10
  • $\begingroup$ Many issues related to polytopes are discussed in Branko Grunbaum's excellent book Convex Polytopes. $\endgroup$ – Joseph Malkevitch Jun 17 '13 at 16:33

This is essentially an issue of semantics. There are two ways of defining polytopes: the V-representation (the convex hull of a finite set of points) and the H-representation (the intersection of half-spaces).

V-representations are always compact, but H-representations don't have to be. A lot of people would reserve the word "polytope" for these objects when they are compact. Can you see how $P$ is the intersection of two halfspaces?

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.