[Paper Review] The Euler characteristic of a category

Tom Leinster|arXiv (Cornell University)|Oct 8, 2006

Homotopy and Cohomology in Algebraic Topology13 references4 citations

TL;DR

This paper introduces a new definition of Euler characteristic for finite categories, generalizing classical notions from topology and combinatorics. It uses a weighted Möbius inversion framework to define the Euler characteristic as the sum of object weights derived from the inverse of the category's adjacency matrix, extending compatibility with orbifolds, groupoids, and topological spaces while providing a unified formula for colimit cardinality in set-valued functors.

ABSTRACT

The Euler characteristic of a finite category is defined and shown to be compatible with Euler characteristics of other types of object, including orbifolds. A formula for the cardinality of the colimit of a diagram of sets is proved, generalizing the classical inclusion-exclusion formula. Both rest on a generalization of Mobius-Rota inversion from posets to categories.

Motivation & Objective

• To define a notion of Euler characteristic for finite categories that generalizes and extends classical definitions from topology and combinatorics.
• To establish compatibility of this Euler characteristic with known invariants such as orbifold invariants, groupoid cardinality, and topological Euler characteristic.
• To provide a general formula for the cardinality of colimits of set-valued functors, generalizing inclusion-exclusion.
• To extend Möbius–Rota inversion from posets to general finite categories, enabling new enumeration and representation techniques.

Proposed method

• Define the Euler characteristic of a finite category as the sum of the weights of its objects, where weights are derived from the inverse of the matrix Z whose (i,j)-entry counts arrows from object i to j.
• Introduce the concept of 'weight' of an object using a generalized Möbius inversion over categories, extending Rota’s theory beyond posets.
• Use the category of elements construction to relate functors to colimits and analyze flatness conditions for representability.
• Establish a representation formula for functors that are sums of representables, using the inverse of the adjacency matrix Z.
• Prove a general formula for the cardinality of a colimit of a set-valued functor, valid for any shape of diagram.
• Demonstrate that the Euler characteristic is compatible with products, fibrations, and other categorical constructions.

Experimental results

Research questions

• RQ1 How can the Euler characteristic of a finite category be defined in a way that generalizes the classical Euler characteristic of topological spaces and orbifolds?
• RQ2 What is the categorical analogue of Möbius inversion, and how does it extend Rota’s theory from posets to general categories?
• RQ3 Can a unified formula for the cardinality of a colimit of a set-valued functor be derived, generalizing inclusion-exclusion?
• RQ4 How does the new Euler characteristic relate to existing invariants such as groupoid cardinality and orbifold invariants?
• RQ5 In what sense is the Euler characteristic a generalized notion of cardinality, and what categorical properties does it satisfy?

Key findings

• The Euler characteristic of a finite category is defined as the sum of the weights of its objects, where weights are derived from the inverse of the adjacency matrix Z, and this definition is rational-valued, not necessarily integral.
• When the adjacency matrix Z is invertible, the Euler characteristic equals the sum of all entries of Z⁻¹, providing a concise computational formula.
• The definition agrees with the topological Euler characteristic when the classifying space of the category has a well-defined Euler characteristic.
• The Euler characteristic is compatible with products of categories, satisfying χ(A × B) = χ(A)χ(B), and behaves well under fibrations and colimits.
• A general formula for the cardinality of a colimit of a set-valued functor is derived, generalizing inclusion-exclusion to arbitrary diagram shapes.
• The theory extends Möbius–Rota inversion to categories, enabling explicit representation of functors as sums of representables via the inverse of the adjacency matrix.