Main Page | Report this Page
Science Forum Index  »  Mathematics Forum  »  Nifty algebra theorem...
Page 1 of 1    

Nifty algebra theorem...

Author Message
Larry Hammick...
Posted: Mon Nov 02, 2009 2:45 am
Guest
Another virtuoso étude from the Sortov Institute:

Let f(A,B,...,K) be a polynomial in any number of variables over Z, and let
n be a positive integer. Define a polynomial F by

F(A,...,K) = sum_{d|n} M(d) f(A^d, B^d, ..., K^d)^{n/d}

where M denotes the Mobius function. Then all the coefficients of F are
divisible by n.

Hints for provers:

Attack first the case
f = A + B + ... + K.
The coefficient of A^a B^b ... K^k, in the polynomial F, turns out to be the
number of mappings g of the cyclic group of n elements into the set
{A,B,...,K}
1) g takes the value A, a times, the value B b times, etc.
2) if g(x) = g(b+x) for all x then b is 0 (mod n).
Since such mapping fall into equivalence classes each with n elements, the
number of those mappings is a multiple of n.

To pass to the general case, add a multiple of n to every coefficient, so
the new coefficients are all positive, and then replace each monomial with a
sum of indeterminates.

Fermat's Little Theorem is the special case in which n is a prime and f is a
constant! Roll over, Fermat.

LH
 
 
Page 1 of 1    
All times are GMT - 5 Hours
The time now is Tue Dec 15, 2009 12:12 am