From: Boudewijn Moonen Subject: Re: betti numbers Date: Fri, 25 May 2001 14:32:15 +0200 Newsgroups: sci.math Summary: Equivalent complexes to determine homology of a CW complex Michael wrote: > > I was wondering if there is a way to compute the betti numbers of a manifold > without knowing the de rham cohomology of the manifold? A supplement to what Robin wrote: Often it is more convenient and economical to use a finite CW-decomposition of a space X in place of a simplicial one. Given such, define the *n-th cellular* chain group C_n(X) as C_n(X) := free abelian group on the n-cells of X and the boundary map del_n : C_n --> C_{n-1}(X) as del_n(e) := sum_{(n-1) - cells e'} [e:e']e' where the *incidence number* [e:e'] of e with e' is defined as follows: Let X_n denote the union of all k-cells with k <= n. Let f : S^{n-1} --> X_{n-1} be the attaching map for e. Then [e:e'] = deg(g), the mapping degree of the map g : S^{n-1} --> X_{n-1} --> X_{n-1}/ X_{n-1}-e' = S^{n-1} the left map being f and the second map collapsing everything in X_{n-1} outside of e' to a point. Then the homology of the chain complex del_n del_{n-1} ... --> C_n(X) ------> C_{n-1}(X) -----------> ... computes the homology, and hence the Betti numbers, of X. This method is quite effective in computing these for the complex and real projective spaces. This all is explained in Dold's Algebraic Topology book. An online reference is Allen Hatcher's book; see http://math.cornell.edu/~hatcher The main difficulty, compared to a simplicial decomposition, is to figure out the incidence numbers. The gain is that cellular decompositions have usually a far lesser number of cells. Regards, -- Boudewijn Moonen Institut fuer Photogrammetrie der Universitaet Bonn Nussallee 15 D-53115 Bonn GERMANY e-mail: Boudewijn.Moonen@ipb.uni-bonn.de Tel.: GERMANY +49-228-732910 Fax.: GERMANY +49-228-732712