A number theory problem by Áron Bán-Szabó

6 ( 6 1 ) ! = 5 × 4 × 3 × 2 × 1 7 ∤ ( 7 1 ) ! = 6 × 5 × 4 × 3 × 2 × 1 8 ( 8 1 ) ! = 7 × 6 × 5 × 4 × 3 × 2 × 1 9 ( 9 1 ) ! = 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 10 ( 10 1 ) ! = 9 × 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 11 ∤ ( 11 1 ) ! = 10 × 9 × 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 12 ( 12 1 ) ! = 11 × 10 × 9 × 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 \begin{aligned} {\color{#D61F06}{6}} & \mid (6-1)!=5\times 4\times 3\times 2\times 1 \\ {\color{#3D99F6}{7}} & \not\mid (7-1)!=6\times 5\times 4\times 3\times 2\times 1 \\ {\color{#D61F06}{8}} & \mid (8-1)!=7\times 6\times 5\times 4\times 3\times 2\times 1 \\ {\color{#D61F06}{9}} & \mid (9-1)!=8\times 7\times 6\times 5\times 4\times 3\times 2\times 1 \\ {\color{#D61F06}{10}} & \mid (10-1)!=9\times 8\times 7\times 6\times 5\times 4\times 3\times 2\times 1 \\ {\color{#3D99F6}{11}} & \not\mid (11-1)!=10\times 9\times 8\times 7\times 6\times 5\times 4\times 3\times 2\times 1 \\ {\color{#D61F06}{12}} & \mid (12-1)!=11\times 10\times 9\times 8\times 7\times 6\times 5\times 4\times 3\times 2\times 1 \end{aligned}

The red numbers are the composite numbers (bigger than 5 5 ), the blue numbers are the prime numbers (bigger than 5 5 ).

Is it true, that for any composite positive integer k k ( > 5 >5 ), k ( k 1 ) ! = ( k 1 ) × ( k 2 ) × × 1 {\color{#D61F06}{k}}\mid (k-1)!=(k-1)\times (k-2)\times \dots \times 1

No, it is false Yes, it is true

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.

2 solutions

Áron Bán-Szabó
Jul 31, 2017

If k 6 k\geq 6 and k k is composite, then k = p q k=pq , where 1 < p < q < k 1 1<p<q<k-1 or k = p 2 k=p^2 , where 1 < p < k 1 1<p<k-1 .

A) k = p q 1 < p < q < k k=pq\quad 1<p<q<k

( k 1 ) ! = 1 × 2 × × p × × q × × ( k 1 ) (k-1)!=1\times 2\times\dots \times p\times \dots\times q\times \dots\times (k-1)

So now the statement is definitly true .

B) k = p 2 1 < p < k p 3 k=p^2\quad 1<p<k\quad p\geq3

It is easy to see that now 2 p < p 2 1 2p<p^2-1 , and

( k 1 ) ! = 1 × 2 × × p × × 2 p × × ( k 1 ) (k-1)!=1\times 2\times\dots\times p\times\dots\times 2p\times\dots\times(k-1)

So the statement is always true .


It's easy to see that if p p is a prime number, then p ∤ ( p 1 ) ! p\not\mid (p-1)! but since Wilson's theorem p ( p 1 ) ! + 1 p\mid (p-1)!+1

Zach Abueg
Jul 31, 2017

Recall the fundamental theorem of arithmetic : every integer > 1 > 1 is either prime or composite, in which case it is the product of a unique combination of smaller primes.

Furthermore, if a number A A divides another number B > A B > A , then all factors of A A will also divide B B . For instance, 10 10 divides 20 20 , so 1 1 , 2 2 , and 5 5 must also divide 20 20 . Any prime p p will only divide a number N > p N > p if multiples of p p also divide N N . This is why p p never divides ( p 1 ) ! (p - 1)! , as seen in the numbers marked in blue.

Now consider a composite number x x with the prime factorization

p 1 q 1 p 2 q 2 p 3 q 3 p k q k \displaystyle p_1^{q_1}p_2^{q_2}p_3^{q_3} \cdots p_k^{q_k}

for primes p 1 , p 2 , p 3 p k < x p_1, p_2, p_3 \ldots p_k < x . Now, it is elementary that x x ! x 0 x \mid x! \ \forall \ x \geq 0 . But is it true that x ( x 1 ) ! x 0 x \mid (x - 1)! \ \forall \ x \geq 0 as well? Yes! , because ( x 1 ) ! (x - 1)! is the product of all positive integers x 1 \leq x - 1 , which must necessarily contain its prime factors.

This solution is incomplete/vague in its current form.

What is missing, is the handling of the powers of primes (multiplicity). E.g. you have to show, that 24! contains the 5 at least twice or 63! the 2 six (or more) times. (The solution of Áron Bán-Szabó shows a way around this by using a neat trick.)

Zee Ell - 3 years, 10 months ago

Log in to reply

Ah, good point! How should I compensate for this? I can see it intuitively but I'm not quite sure how to show it rigorously.

Zach Abueg - 3 years, 10 months ago

Log in to reply

If you could prove, that p n ( p n 1 ) ! , that would do the trick. \text {If you could prove, that } p^n | (p^n - 1)! \text { , that would do the trick.}

Tip: it is enough to show, that n × p p n 1 for n > 5 . \text {Tip: it is enough to show, that } n×p ≤ p^n - 1 \text { for n > 5 .}

Zee Ell - 3 years, 10 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...