3
$\begingroup$

In dimension 2, the euler poincare formula restricts the incidence properties of edges in a triangulation of a surface. Are there analogous generalizations for higher dimensions, like elaborations on the Schenzel formula for the specific case pf spheres?

$\endgroup$
1
  • $\begingroup$ For a triangulated manifold of dimension $d > 0$, the $(d-1)$-dimensional simplices must all have exactly two $d$-dimensional co-faces. $\endgroup$ Dec 7, 2015 at 17:42

2 Answers 2

4
$\begingroup$

The $g$-conjecture (or $g$-theorem in the polytopal case) aims the characterize what possible $f$-vectors can arise from simplicial spheres. Here is some nice history and introduction to the $g$-conjecture by Gil Kalai. In brief the $g$-conjecture says that for a simplicial sphere the $h$-vector should be symmetric and the $g$-vector should be an $M$-vector where $g_i = h_i - h_{i-1}$.

$\endgroup$
0
3
$\begingroup$

As is alluded to in the article linked in the previous answer, the Dehn-Sommerville equations are a powerful generalization of the Euler-Poincare formula for a homology sphere. These say that, for a homology $d$-sphere, the relation $h_i = h_{d-i}$ holds for every $i \in {0,\dots, d}$. This simple $h$-vector relation can be expanded out into a rather more complicated (but equivalent) relation on the $f$-vector. In particular, the relation $h_d = h_0$ recovers the Euler-Poincare formula.

There's a nice Wikipedia article on the Dehn-Sommerville equations. In particular, you can look here if you want to see the $f$-vector formulation of the equations.

The Wikipedia article only mentions simplicial polytopes. If you want a reference for the statement on homology spheres, see the Kalai article (informally), or, in somewhat different language, Stanley's book Combinatorics and Commutative Algebra.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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