Let be a function from the set of positive integers to the set of positive integers such that for all positive integers , and for all positive integers . Find the number of possible values of .
This problem is from the OMO.
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.
Throughout this solution, f n ( x ) = f ( f ( . . . f ( x ) ) . . . ) , where the function f is applied n times.
f is a function from the set of positive integers to itself such that
( 1 ) f ( x ) ≤ x 2
( 2 ) f ( f 2 ( x ) f 2 ( y ) ) = x y
for all positive integers x , y .
( 1 ) implies f ( 1 ) = 1 . Substituing y = 1 in ( 2 ) together with f ( 1 ) = 1 gives f 3 ( x ) = x for all positive integers x .
Step 1. f is bijective
If f ( x ) = f ( y ) , then x = f 3 ( x ) = f 3 ( y ) = y , so f is injective. Since f 3 ( x ) = x , for every x there exists y = f 2 ( x ) such that f ( y ) = x , so f is surjective. Thus, f is bijective.
Step 1 implies f − 1 = f 2 is defined.
Step 2. f ( x ) f ( y ) = f ( x y ) and f − 1 ( x ) f − 1 ( y ) = f − 1 ( x y )
By ( 2 ) , f ( f − 1 ( x ) f − 1 ( y ) ) = f ( f 2 ( x ) f 2 ( y ) ) = x y . Thus, by applying f − 1 to both sides, we get the second identity. Similarly, f ( x ) f ( y ) = f ( f − 1 ( f ( x ) f ( y ) ) ) = f ( f − 1 ( f ( x ) ) f − 1 ( f ( y ) ) ) = f ( x y ) .
Step 3. If p is prime, then f ( p ) is prime.
Suppose f ( p ) = q r for some prime p and some product q r with q , r ≥ 2 . Then, p = f − 1 ( q r ) = f − 1 ( q ) f − 1 ( r ) , so p is the product of two numbers greater than 1 , a contradiction. (Since f ( 1 ) = 1 and f is bijective.)
Thus, we can describe all possible f as a set of ordered triples of primes ( p , q , r ) , such that f ( p ) = q , f ( q ) = r , f ( r ) = p , (WLOG, p < q , r ), and with every prime appearing in at most one triple and for all triples p 4 ≥ q 2 ≥ r . If a prime s appears in no triples, then f ( s ) = s .
Step 4. All such sets of triples describe a valid f .
Consider any x = ∏ p i where each p i is prime. Then, f ( x ) = f ( ∏ p i ) = ∏ f ( p i ) ≤ ∏ p i 2 = x 2 (by ( 1 ) ). So, f satisfies ( 1 ) . Let y = ∏ q i where each q i is prime. Then, f ( x ) f ( y ) = ∏ f ( p i ) ∏ f ( q i ) = f ( x y ) (because p i and q i together are the prime factors of x y .) Thus, f satisfies ( 2 ) .
Now, consider f ( 3 0 ) = f ( 2 ) f ( 3 ) f ( 5 ) . We know that f ( 2 ) ∈ { 2 , 3 } . If f ( 2 ) = 3 , then f ( 3 ) ∈ { 5 , 7 } . If f ( 3 ) = 5 , then f ( 5 ) = 2 , giving f ( 3 0 ) = 3 0 ; if f ( 3 ) = 7 , then f ( 5 ) ∈ { 5 , 1 1 , 1 3 , 1 7 , 1 9 , 2 3 } , each of them giving a new value of f ( 3 0 ) . Now, we have 7 values, all of which are odd except for 3 0 .
If f ( 2 ) = 2 , then f ( 3 ) ∈ { 3 , 5 , 7 } . Either way, f ( 5 ) ∈ { 5 , 7 , 1 1 , 1 3 , 1 7 , 1 9 , 2 3 } . Except
a) f ( 5 ) = 5 if f ( 3 ) = 5 .
b) f ( 5 ) = 7 if f ( 3 ) = 7 .
c) If f ( 3 ) = 3 and f ( 5 ) = 5 , then f ( 3 0 ) = 3 0 , which we already counted.
d) We count f ( 3 0 ) = 2 ⋅ 5 ⋅ 7 twice.
That's gives a total of 7 ⋅ 3 − 4 = 1 7 more values, all of which are even and not 3 0 .
This gives a total of 7 + 1 7 = 2 4 possible values.