Factor A Degree 100000000 Polynomial

Algebra Level 5

We are considering polynomials of the form ( x 1 ) ( x 2 ) ( x 3 ) ( x p + 1 ) + 1 p 2 (x-1)(x-2)(x-3) \cdots (x-p +1) + 1 -p^2 that can be factorized into 2 non-constant polynomials with integer coefficients.

What is the number of those polynomials if p is prime and less than 1 0 8 10^8 ?


Inspiration .

Between 100 and 999 Between 10 and 99 Between 1 and 9 Equal to 0 Between 1000 and 9999 More than 9999

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

1 solution

Mark Hennings
Nov 22, 2016

If p p is prime, then a j 1 a^j \equiv 1 for all 1 a p 1 1 \le a \le p-1 , so that x p 1 1 j = 1 p 1 ( x j ) x^{p-1} - 1 \; \equiv \; \prod_{j=1}^{p-1}(x - j) in Z p [ x ] \mathbb{Z}_p[x] . Thus, if f p ( x ) = j = 1 p 1 ( x j ) + 1 p 2 f_p(x) \; = \; \prod_{j=1}^{p-1}(x-j) + 1 - p^2 then f p ( x ) x p 1 f_p(x) \,\equiv\, x^{p-1} in Z p [ x ] \mathbb{Z}_p[x] . If p 2 p^2 fails to divide ( p 1 ) ! + 1 (p-1)!+1 , then p 2 p^2 does not divide the constant coefficient of f p ( x ) f_p(x) , and hence f p ( x ) f_p(x) is irreducible in Z [ x ] \mathbb{Z}[x] by Eisenstein's irreducibility criterion.

Thus the primes p p for which Eisenstein's Criterion, using the prime p p , fails to show that f p ( x ) f_p(x) is irreducible are the ones for which p 2 p^2 divides ( p 1 ) ! + 1 (p-1)!+1 . These primes are called Wilson primes . Only three Wilson primes are known up to 5 × 1 0 8 5 \times 10^8 , these being 5 5 , 13 13 and 563 563 . Since f 5 ( x ) f_5(x) has no constant term, it certainly is reducible. I haven't checked the reducibility of f 13 ( x ) f_{13}(x) and f 563 ( x ) f_{563}(x) , but it is clear that the number of primes p p between 1 1 and 1 0 8 10^8 for which f p ( x ) f_p(x) is reducible is at least 1 1 and no more than 3 3 , making between 1 and 9 \boxed{\mbox{ between } 1 \mbox{ and } 9} the correct answer.

This is a marvelous problem and solution!

I'm amazed how far that one problem has inspired you. Would love to see more :)

Calvin Lin Staff - 4 years, 6 months ago

Log in to reply

This is no limit to inspiration.
The only constraints I can see are:
to not become too complex
and to have the answer to the problem

Abdelhamid Saadi - 4 years, 6 months ago

Short and sweet.

I wonder if there is a simple way to check the reducibility of f 13 f_{13} and f 563 f_{563} .

Abdelhamid Saadi - 4 years, 6 months ago

Log in to reply

On computer checking with Mathematica, it seems that both f 13 f_{13} and f 563 f_{563} are irreducible. I have yet to find a good reason why. It seems that f 13 f_{13} is irreducible modulo 11 11 ; if that were easyish to prove, that would confirm the irreducibility of f 13 f_{13} . There is no prime p p less than 563 563 for which f 563 f_{563} becomes irreducible modulo p p , so the modulus-reduction technique will not help in this case.

It seems, however, that p = 5 p=5 is the only case where f p f_p is reducible for p < 5 × 1 0 8 p < 5 \times 10^8 .

Mark Hennings - 4 years, 6 months ago

Log in to reply

May be we should consider f p , k ( x ) = ( x 1 ) ( x 2 ) . . ( x p + 1 ) + 1 + k p 2 f_{p,k}(x) = (x - 1)(x-2)..(x-p+1) + 1 + kp^{2} for k in Z \mathbb{Z}

The proof will be the same. only f 5 , k , f 13 , k , f 563 , k f_{5,k}, f_{13,k}, f_{563,k} could be reducible.

If we take k = ( p 1 ) ! + 1 p 2 k = - \dfrac {(p-1)! + 1}{p^2} , it will be reducible.

Abdelhamid Saadi - 4 years, 6 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...