Computing homotopy classes for diagrams

TL;DR

This paper introduces a polynomial-time algorithm for computing equivariant homotopy classes of maps between finite simplicial sets under certain stability conditions, with applications to computational topology and Tverberg-type problems.

Contribution

It provides the first polynomial-time algorithm for equivariant homotopy class computation in the stable range, extending to diagrams and applications in topological combinatorics.

Findings

Algorithm computes equivariant homotopy classes efficiently

Applicable to compute equivariant stable homotopy groups

Decides Tverberg-type intersection problems in polynomial time

Abstract

We present an algorithm that, given finite simplicial sets $X$, $A$, $Y$ with an action of a finite group $G$, computes the set $[X,Y]^A_G$ of homotopy classes of equivariant maps $\ell \colon X \to Y$ extending a given equivariant map $f \colon A \to Y$ under the stability assumption $\dim X^H \leq 2 \operatorname{conn} Y^H$ and $\operatorname{conn} Y^H \geq 1$, for all subgroups $H\leq G$. For fixed $n = \operatorname{dim} X$, the algorithm runs in polynomial time. When the stability condition is dropped, the problem is undecidable already in the non-equivariant setting. The algorithm is obtained as a special case of a more general result: For finite diagrams of simplicial sets $X$, $A$, $Y$, i.e. functors $\mathcal{I}^\mathrm{op} \to \mathsf{sSet}$, in the stable range $\operatorname{dim} X \leq 2 \operatorname{conn} Y$ and $\operatorname{conn} Y > 1$, we give an algorithm that computes the set $[X, Y]^A$ of homotopy classes of maps of diagrams $\ell \colon X \to Y$ extending a given $f \colon A \to Y$. Again, for fixed $n = \dim X$, the running time of the algorithm is polynomial. The algorithm can be utilized to compute homotopy invariants in the equivariant setting -- for example, one can algorithmically compute equivariant stable homotopy groups. Further, one can apply the result to solve problems from computational topology, which we showcase on the following Tverberg-type problem: Given a $k$-dimensional simplicial complex $K$, is there a map $K \to \mathbb{R}^{d}$ without $r$-tuple intersection points? In the metastable range of dimensions, $rd \geq (r+1)k +3$, the result of Mabillard and Wagner shows this problem equivalent to the existence of a particular equivariant map. In this range, our algorithm is applicable and, thus, the $r$-Tverberg problem is algorithmically decidable (in polynomial time when $k$, $d$ and $r$ are fixed).