lunes, 20 de febrero de 2012

Abelian Group

A group for which the elements commute (i.e., AB=BA for all elements 
A and B) is called an Abelian group. 

All cyclic groups are Abelian, but an Abelian group is not necessarily cyclic. 

All subgroups of an Abelian group are normal. 

In an Abelian group, each element is in a conjugacy class by itself, and the character table involves powers of a single element known as a group generator.

In Mathematica, the function AbelianGroup[{n1, n2, ...}] represents the direct product of the cyclic groups of degrees n_1n_2, ....

No general formula is known for giving the number of nonisomorphic
 finite groups of a given group order. 

However, the number of nonisomorphic Abelian finite groups a(n) of any given group order n is given by writing n as

 n=product_(i)p_i^(alpha_i),
where the p_i are distinct prime factors, then


 a(n)=product_(i)P(alpha_i),
(2)
where P(k) is the partition function, which is implemented 
in Mathematica as FiniteAbelianGroupCount[n].

 The values of a(n) for n=1, 2, ... are 1, 1, 1, 2, 1, 1, 1, 3, 2, ... 

(Sloane's A000688).

The smallest orders for which n=1, 2, 3, ... nonisomorphic Abelian groups exist are 1, 4, 8, 36, 16, 72, 32, 900, 216, 144, 64, 1800, 0, 288, 128, ...

 (Sloane's A046056), where 0 denotes an impossible number

(i.e., not a product of partition numbers) of nonisomorphic Abelian, groups.

 The "missing" values are 13, 17, 19, 23, 26, 29, 31, 34, 37, 38, 39, 41, 43, 46, ... (Sloane's A046064). 

The incrementally largest numbers of Abelian groups as a function of order are 1, 2, 3, 5, 7, 11, 15, 22, 30, 42, 56, 77, 101, ... 

(Sloane's A046054), which occur for orders 1, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192, ... (Sloane's A046055).

The Kronecker decomposition theorem states that every finite Abelian group can be written as a group direct product of cyclic groups of prime power
 group order.

 If the group order of a finite group is a prime p
then there exists a single Abelian group of order p (denoted Z_p
and no non-Abelian groups.

 If the group order is a prime squared p^2, then there are two Abelian groups (denoted Z_(p^2) and Z_p×Z_p. If the group order is a prime cubed p^3, then there are three Abelian groups (denoted Z_p×Z_p×Z_pZ_p×Z_(p^2), and Z_(p^3)), and five groups total. If the order is a productof two primes p and q, then there exists exactly one Abelian group of group order pq (denoted Z_p×Z_q).

Another interesting result is that if a(n) denotes the number of nonisomorphic Abelian groups of group order n, then

 sum_(n=1)^inftya(n)n^(-s)=zeta(s)zeta(2s)zeta(3s)...,
(3)
where zeta(s) is the Riemann zeta function.

The numbers of Abelian groups of orders <=n are given by 1, 2, 3, 5, 6, 7, 8, 11, 13, 14, 15, 17, 18, 19, 20, 25, ... (Sloane's A063966) 

for n=1, 2, .... Srinivasan (1973) has also shown that

 sum_(n=1)^Na(n)=A_1N+A_2N^(1/2)+A_3N^(1/3)+O[N^(105/407)(lnN)^2],
(4)

where

A_k=product_(j=1; j!=k)^(infty)zeta(j/k)
(5)
={2.294856591... for k=1; -14.6475663... for k=2; 118.6924619... for k=3,
(6)

(Sloane's A021002, A084892, and A084893) and zeta(s) is again the 
Riemann zeta function. Note that Richert (1952) incorrectly gave A_3=114

The sums A_k can also be written in the explicit forms

A_1=product_(j=2)^(infty)zeta(j)
(7)
A_2=zeta(1/2)product_(j=3)^(infty)zeta(1/2j)
(8)
A_3=zeta(1/3)zeta(2/3)product_(j=4)^(infty)zeta(1/3j).
(9)

DeKoninck and Ivic (1980) showed that

 sum_(n=1)^N1/(a(n))=BN+O[sqrt(N)(lnN)^(-1/2)],
(10)
where

B=product_(p){1-sum_(k=2)^(infty)[1/(P(k-1))-1/(P(k))]1/(p^k)}
(11)

=
0.752...
(12)

(Sloane's A084911) is a product over primes p and P(n) is again the partition function.

No hay comentarios: