San José State University

applet-magic.com
Thayer Watkins
Silicon Valley
USA

The Euler-Poincaré Formula
in Algebraic Topology

Euler's Formula for Polyhedra

The regular polyhedra were known at least since the time of the ancient Greeks. The names of the more complex ones are purely Greek. But despite their being known for close to two millenia no one apparently noticed the fact that the sum of the number of faces F and the number of vertices V less the number of edges E is equal to two for all of them; i.e.,


V - E + F = 2
 

It was the Swiss mathematician Leonhard Euler who recognized and published this fact. The value of two is said to be the Euler characteristic of each of the polyhedra. This value is not changed by stretching or shrinking any side or face or even shrinking a side or face to zero. This means that the Euler characteric is a topological invariant because it is not altered by any continuous mapping.

The quantity (V-E+F) is called the Euler characteristic of the polyhedron and is denoted as by the Greek letter Chi, Χ. Thus Χ(cube)=2.

The Generalization of Euler's Formula

The French-Swiss mathematician, Simon Lhuilier (1750-1840), found a slight generalization of Euler's formula to take into account polyhedra having holes. Lhuilier's formula is


V - E + F = 2 − 2G = 2(1− G)
 

where G is the number of holes in the polyhedron. Thus the Euler characteristic is 2 for a regular polyhedron but 0 for a torus-like polyhedron.

For a simple treatment of the effect of holes and handles on the Euler characteristic see Euler Characteristic.

It was the French mathematician, Henri Poincaré, who fully generalized Euler's formula.

Consider an n-dimensional polytope K. It is made up of simplexes of equal or lesser dimensions. A two dimensional simplex (2-simplex) is a triangle; a one dimensional simplex (1-simplex) is a line and a zero dimensional simplex (0-simplex) is a point. Let αm be the number of m-dimensional simplexes in K, where m runs from 0 to n. For example, the surface of a tetrahedron is two dimensional and it is composed of 4 triangles (2-simplexes), 6 line segment edges (1-simplexes) and 4 points (0-simplexes).

Chains

Consider a sequential labeling of the m-simplexes of K taking into account their orientation; i.e., choose an m-simplex with a particular orientation and call it number 1, another as number 2 and so on. Now construct vectors with integer components of the form


C = (c1, c2, ..., cαm)
 

where the ci's are integers (ci∈Z),

One of these vectors is called a chain and the set of them form a mathematical group under the operation of component-wise addition with the identity being the vector of zeroes and the inverse of a vector being the vector of the negatives of the components. The group is abelian and its dimension is αm.

A chain may be represented as a formal sum of the form


C = c1σ1m + c2σ2m + ... + cαmσαmm
 

where σim is the unit vector for the i-th simplex; i.e.,


σim = (0,0,..,1,..,0,0) with the "1" in the i-th place.
 

This is essentially the same as representing a vector in 3-space as as the formal sum


V = xi + yj + zk
 

where i, j and k are the unit vectors in the three directions and equivalent to (1,0,0), (0,1,0) and (0,0,1).

The Boundary Operator ∂

The boundary for σim is defined as


∂(σim) = Σ [σim, σjm-1im
 

where [σim, σjm-1] is the incidence number of the σim simplex with the σjm-1 simplex; i.e.,


im, σjm-1] = 0 if σjm-1 is not a face of σim
im, σjm-1] = ±1 if σjm-1 is a face of σim
 

with the sign of the incidence number depending upon the orientation of σjm-1 with respect to the orientation of σim.

The boundary of C = Σciσjm is then


∂C = Σci∂(σjm)
= ΣiciΣjim, σjm-1jm-1
 

The Boundary of a Boundary

Because of a property of the incidence numbers that for all i and k


Σ [σim, σjm-1] [σjm-1, σkm-2] = 0
where the sum is over all values of j.
 

it follows that


∂(∂C) = 0
the boundary of a boundary is empty.
 

In other words, boundaries have no boundary.

Cycles and m-Boundaries

The boundary operation defines a homomorphism of the group of chains on m-simplexes, Cm(K) onto the group of chains on the (m-1)-simplexes, Cm-1(K):


∂:Cm(K)→Cm-1(K).
 

Within the group of chains on m-simplexes there are special chains that have an empty boundary; i.e. ∂(C)=0. These boundary-less chains are called m-cycles and they form a subgroup of Cm(K) denoted as Zm(K).

Within the group of m-cycles there are the special cycles that are the boundary of some chain in Cm+1(K). They form a subgroup of Zm(K) and hence a subgroup of Cm(K). They are called m-boundaries and their group is denoted as Bm(K).

Homomorphism of Groups

The subgroup relationship of Cm(K), Zm(K) and Bm(K) determines a factor group of Zm(K) with respect to Bm(K). This is called the m-th homology group of the polytope K and is denoted as Hm(K)

The boundary operator ∂ maps Cm(K) onto Bm-1(K)(K) and it maps Zm(K) onto the identity element of Bm-1(K), which is also the identity element of Cm-1(K) and Zm-1(K). Zm(K) is said to be the kernal of the homomorphism. This means that Zm(K) is a factor group of Cm(K) with respect to Bm-1(K); i.e.,


Zm(K) = Cm(K)/Bm-1(K)
or, in additive notation
Zm(K) = Cm(K) - Bm-1(K)
 

From the theorems for abelian free groups it follows then for m>0


rank(Zm(K)) = rank(Cm(K)) - rank(Bm-1(K)
 

Letting rank(Zm(K))=ζm, rank(Cm(K))=αm, and rank(Bm(K))=βm the above condition is


ζm = αm - βm-1
 

For the special case of m=0, rank(Bm-1(K))=0 so


ζ0 = α0
 

Let rank(Hm(K))=pm. The number pm is known as the m-th Betti number of K, so named after the Italian mathematician Enrico Betti (1923-1892) who believed that the structure of a topological space could be completely characterized by a set of numerical invariants. Now the characterization of a topological space is in terms of the topologically invariant groups Hm(K). Since Hm(K)=Zm(K)-Bm(K) then


pm = ζm - βm
in particular
p0 = ζ0 - β0 c
which, since β0=0, reduces to
p0 = ζ0
 

The relevant equations for m>0 placed together are:


αm = ζm - βm-1
pm = ζm - βm
 

Subtracting the second equation from the first gives


αm - pm = βm-1 + βm
for m>0
and
α0 - p0 = 0
 

The Euler-Poincaré Formula

Consider now the alternating sum


Σnm=0(-1)mm-pm) = Σnm=0(-1)mm-1 + βm)
 

The right hand side of the above reduces to


0 - β1 - β0 + β2 + β1
which collapses to
(-1)nβn
 

Since Cn+1(K)=0 it follows that Bn(K)=0 and thus βn=0 so


Σnm=0(-1)mm-pm) = 0
and hence
Σnm=0(-1)mαm = Σnm=0(-1)mpm
 

This latter equation is the Euler-Poincaré Formula. The left hand side is the Euler characteristic of the polytope K.


HOME PAGE OF applet-magic
HOME PAGE OF Thayer Watkins