nLab
groupoid cardinality

Context

Homotopy theory

Higher category theory

higher category theory

Basic concepts

Basic theorems

Applications

Models

Morphisms

Functors

Universal constructions

Extra properties and structure

1-categorical presentations

Contents

Idea

The homotopy cardinality or -groupoid cardinality of a (sufficiently “finite”) space or ∞-groupoid X is an invariant of X (a value assigned to its equivalence class) that generalizes the cardinality of a set (a 0-truncated -groupoid).

Specifically, whereas cardinality counts elements in a set, the homotopy cardinality counts objects up to equivalences, up to 2-equivalences, up to 3-equivalence, and so on.

This is closely related to the notion of Euler characteristic of a space or -groupoid. See there for more details.

Definition

Groupoid cardinality

The cardinality of a groupoid X is the real number

X= [x]π 0(X)1Aut(x),|X| = \sum_{[x] \in \pi_0(X)} \frac{1}{|Aut(x)|} \,,

where the sum is over isomorphism classes of objects of X and Aut(x) is the cardinality of the automorphism group of an object x in X.

If this sum diverges, we say X=. If the sum converges, we say X is tame.

-Groupoid cardinality

This is the special case of a more general definition:

The groupoid cardinality of an ∞-groupoid X – equivalently the Euler characteristic of a topological space X (that’s the same, due to the homotopy hypothesis) – is, if it converges, the alternating product of cardinalities of the (simplicial) homotopy groups

X:= [x]π 0(X) k=1 π k(X,x) (1) k= [x]1π 1(X,x)π 2(X,x)1π 3(X,x)π 4(X,x).|X| := \sum_{[x] \in \pi_0(X)}\prod_{k = 1}^\infty |\pi_k(X,x)|^{(-1)^k} = \sum_{[x]} \frac{1}{|\pi_1(X,x)|} |\pi_2(X,x)| \frac{1}{|\pi_3(X,x)|} |\pi_4(X,x)| \cdots \,.

This corresponds to what is referred to as the total homotopy order of a space, introduced by Quinn in notes in 1995 on TQFTs (see reference list).

Examples

  • Let X be a discrete groupoid on a finite set S with n elements. Then the groupoid cardinality of X is just the ordinary cardinality of the set S

    X=n.|X| = n \,.
  • Let BG be the delooping of a finite group G with k elements. Then

    BG=1k|\mathbf{B}G| = \frac{1}{k}
  • Let A be an abelian group with k elements. Then we can deloop arbitrarily often and obtain the Eilenberg–Mac Lane objects B nA for all n. (Under the Dold–Kan correspondence B nA is the chain complex A[n] (or A[n] depending on notational convention) that is concentrated in degree n, where it is the group A). Then

    B nA={k ifniseven 1k ifnisodd|\mathbf{B}^n A| = \left\{ \array{ k & if n is \; even \\ \frac{1}{k} & if n is \; odd } \right.
  • Let E=core(FinSet) be the groupoid of finite sets and bijections – the core of FinSet. Its groupoid cardinality is the Euler number

    E= n1S n= n1n!=e.|E| = \sum_{n\in \mathbb{N}} \frac{1}{|S_n|} = \sum_{n\in \mathbb{N}} \frac{1}{n!} = e \,.
  • Let E=(E i) be a finite crossed complex, (i.e., an omega-groupoid; see the work of Brown and Higgins) such that for any object vE 0 of E the cardinality of the set of i-cells with source v is independent of the vertex v. Then the groupoid cardinality of E can be calculated as E= i#(E i) (1) i, much like a usual Euler characteristic. For the case when F is a totally free crossed complex, this gives a very neat formula for the groupoid cardinality of the internal hom HOM(F,E), in the category of omega-groupoids. Therefore the groupoid cardinality of the function spaces (represented themselves by internal homs) can easily be dealt with if the underlying target space is represented by a omega-groupoid, i.e., has trivial Whitehead products. (This is explored in the papers by Faria Martins and Porter mentioned in the reference list, below.)

References

Revised on November 9, 2012 14:42:10 by Urs Schreiber (82.169.65.155)