For any prime p > 7 , consider the remainders when the fifth powers (i.e. n 5 ) are divided by p . Let f ( p ) be the smallest possible remainder which is not 0 or 1.
The sum of the two smallest possible ratios f ( p ) p can be written as b a , where a and b are coprime positive integers. What are the last three digits of the value of a + b ?
Details and assumptions
As an explicit example, for p = 1 3 , you can verify that the possible remainders are 0 , 1 , 2 , … , 1 2 . Thus f ( 1 3 ) = 2 .
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.
Great job! Nice solution and explanation.
Looks solid!
I have trouble understanding one step; How does a g c d ( 5 , p − 1 ) ≡ b g c d ( 5 , p − 1 ) follow from the previous statements?
Log in to reply
Apply [Bezout's Lemma]: g cd ( x , y ) = n x + m y for some integers n , m .
This is a famous lemma(theorem?).
The smallest case (p=11) can be proven with Fermat's Little Theorem:
a 1 0 ≡ 1 m o d 1 1
Therefore a 5 ≡ 1 or − 1 m o d 1 1 , so f ( 1 1 ) = 1 0 .
I tried to find similar logic for the other cases 1 0 k + 1 but couldn't get anything conclusive and had to fall back on brute force.
To clarify, by "direct computation" you mean that you worked out n^5 mod 101 for all values 2 <= n < 101, etc. ?
Log in to reply
Not necessarily. If we want to calculate the exact value of f ( 1 0 1 ) , then yes we have to check the different values.
If however, we already suspect that 71 leads to the second smallest value, we merely need to check that f ( 1 0 1 ) 1 0 1 > 2 0 ) 7 1 . This can be verified since 4 5 ≡ 1 4 ( m o d 1 0 1 ) , hence f ( 1 0 1 ) ≤ 1 4 .
Similarly, we didn't calculate the exact values of f ( 3 1 ) , f ( 4 1 ) , f ( 6 1 ) , but just showed that f ( p ) p > 4 > 2 0 7 1 .
If p-1 is relatively prime to 5, f(p) is always 2, so the best we can do is 13/2.
Now suppose p-1 is divisible by 5, i.e. p = 10k + 1. Direct computation for k = 1, 3, 4, 6, 7, and 10 gives that the best two fractions are 11/10 and 71/20, which are both less than 13/2. There is also no need to continue further, since 2^5 is always a fifth power and 71/20 < 131/32 <= p/f(p) for p > 101.
Great job! Nice solution, cleanly written.
Mathematically, this solution is the same as Zi Song Y.'s, so if anyone is looking for more details, read the Zi Song Y.'s post.
Excellent!
Problem Loading...
Note Loading...
Set Loading...
FIrstly, if 5 ∤ p − 1 , then g cd ( 5 , p − 1 ) = 1 . We claim that a 5 ≡ b 5 ( m o d p ) ⇒ a ≡ b m o d p . The case when p ∣ a is obvious, so assume that a and b are relatively prime to p . Note that by Fermat's little theorem, a p − 1 ≡ b p − 1 ≡ 1 ( m o d p ) . So, a g cd ( 5 , p − 1 ) ≡ b g cd ( 5 , p − 1 ) ( m o d p ) , which implies a ≡ b ( m o d p ) . Thus, by the claim, there exists an integer n such that n 5 ≡ 2 ( m o d p ) . (For { 1 5 , 2 5 , . . . , ( p − 1 ) 5 } = { 1 , 2 , . . . , p − 1 } mod p ). This implies f ( p ) = 2 . So, the minimum possible value of the fraction f ( p ) p purely depends on p . Thus, the minimum value of the fraction for this case is 2 1 3 . (Since p > 7 .)
Secondly, let 5 ∣ p − 1 . Then, p = 1 0 k + 1 for some integer k . By direct computation for the cases k = 1 , 3 , 4 , 6 , 7 and 1 0 , we find that the two smallest possible values of the fraction is 1 0 1 1 and 2 0 7 1 (obtained from k = 1 , 7 .) . It is clear that 2 0 7 1 < 2 1 3 , so we just have to check that there are no more smaller fractions for k > 1 0 . Since for k = 1 1 , p is not a prime, we just have to check for k ≥ 1 2 or p ≥ 1 2 1 . For p ≥ 1 2 1 , 2 5 ≡ 3 2 ( m o d p ) . So, f ( p ) ≤ 3 2 . Thus, f ( p ) p ≥ 3 2 p ≥ 3 2 1 2 1 ≥ 2 0 7 1 . This implies that the two smallest possible value of the fraction is indeed 1 0 1 1 and 2 0 7 1 . Their sum is 2 0 9 3 . So, a + b = 1 1 3 .
Note : f ( 1 1 ) = 1 0 , f ( 3 1 ) = 5 , f ( 4 1 ) = 3 , f ( 6 1 ) = 1 1 , f ( 7 1 ) = 2 0 , f ( 1 0 1 ) = 6 .