For positive integers x , let f ( x ) be the number of ordered pairs of positive integers ( a , b ) such that a 1 + b 1 = x 1 .
Find the smallest possible value of N such that N is a prime power and f ( N ) = 1 3 .
This problem is posed by Sreejato B.
Details and assumptions
A prime power is of the form p n , where p is a prime number and n is a positive integer.
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.
good Mr. Sotiri,,,best and lucid method!!!
Your solution is very well written. I have a small suggestion though. Use more paragraphs.
I know it takes up more space but it makes solutions a bit more readable.
Instead of factoring, I used the fact that b(p^n) is divisble by b-(p^n) and then used modular arithmetic to know that (p^2n) is divisible by b-(p^n). Rest solution was the same as that of your method.
I wrote a Pascal programming code for this issue :-p
I solved this problem with the assumption that f ( N ) = 1 3 . I shall explain why later. The solutions to the Diophantine equation a 1 + b 1 = x 1 are related to the divisors of x 2 . This is because
a 1 + b 1 = x 1
a b a + b = x 1
From this we have
a b − a x − b x = 0
Adding x 2 to both sides
x 2 + a b − a x − b x = x 2
Factorizing we get
( a − x ) ( b − x ) = x 2
Hence we see that the number of solutions to a 1 + b 1 = x 1 is τ ( x 2 ) which means that f ( x ) = τ ( x 2 ) ,where τ ( x ) is a function that counts the number of divisors of x . It is also well known that for an integer N with prime factorization N = p 1 a 1 . p 2 a 2 . p 3 a 3 . . . τ ( N ) = ( a 1 + 1 ) . ( a 2 + 1 ) . ( a 3 + 1 ) . . .
Now we have:
1 3 = τ ( N 2 )
1 3 = ( a 1 + 1 ) . ( a 2 + 2 ) . ( a 3 + 3 ) . . . = τ ( N 2 ) = f ( N )
but since N is a prime power its factorization should be of the form N = p 1 a 1 + 1
so:
a 1 + 1 = 1 3
a 1 = 1 2
N 2 = p 1 2
N = p 6
If N had been 1 2 we would have got 6.5 as an exponent.
Since we are asked for the smallest possible value of N we take the smallest possible prime 2
Giving us N = 2 6 = 6 4
I wrote a quick hack to generate the pairs ( a , b ) . It revealed that there are 13 of them.So there is no way that f ( N ) = 1 2 . Please correct the question to read f ( N ) = 1 3 .
( 4 1 6 0 , 6 5 ) ( 8 0 , 3 2 0 ) ( 1 9 2 , 9 6 ) ( 3 2 0 , 8 0 ) ( 7 2 , 5 7 6 ) ( 6 6 , 2 1 1 2 ) ( 9 6 , 1 9 2 ) ( 1 2 8 , 1 2 8 ) ( 1 0 8 8 , 6 8 ) ( 6 5 , 4 1 6 0 ) ( 5 7 6 , 7 2 ) ( 2 1 1 2 , 6 6 ) ( 6 8 , 1 0 8 8 )
EDIT: It has been corrected.
Log in to reply
(a,b) is an ordered pair, so for (4160,65) and (65,4160) isn't it the same?
Log in to reply
Actually no,ordered pairs with interchanged x and y values are distinct form each other. If that wasn't the case we would have 6 solutions.
Correct. I too had spent too much time thinking, because of that mistake... I thought to request of dispute, but when mathematicians like Calvin are question setters, it's better to review your argument!
Log in to reply
It is! I think this question might have made a good computer science problem too,building an efficient factorizing function is very tricky,
I like this one!
Best solution! And God bless Ethiopia - I remember their exhibit at Expo 67 in Montreal!
We want to find the number of ordered pairs of positive integers ( a , b ) such that
a 1 + b 1 = N 1 .
Obviously, a , b > N , because a and b are both positive. Let b = N + d for some positive integer d . The equation then becomes
a 1 + N + d 1 = N 1 ⟹ a 1 = N 1 − N + d 1 = N ( N + d ) d .
So,
a = d N ( N + d ) = N + d N 2 .
The number of ordered pairs ( a , b ) that satisfy the equation is therefore the same as the number of values d can take, i.e. the number of positive divisors of N 2 . We know that N is a prime power, i.e.
N = p n ⟹ N 2 = p 2 n ,
and p 2 n has 2 n + 1 divisors. So,
2 n + 1 = 1 3 ⟹ n = 6 ⟹ N = p n = p 6 .
Clearly, the smallest possible value of N = p 6 = 2 6 = 6 4 .
The question should read f ( N ) = 1 3 . If a , b are positive integer solutions of a − 1 + b − 1 = N − 1 , then a , b > N and N b + N a = a b , so ( a − N ) ( b − N ) = N 2 . Thus the positive integer solutions of the equation a − 1 + b − 1 = N − 1 are a = N + α , b = N + β where α , β are positive integer factors of N 2 and α β = N 2 . Thus f ( N ) = d ( N 2 ) is the number of positive divisors of N 2 , which is always odd. For N = p n and f ( N ) = 1 3 , we need 2 n + 1 = d ( p 2 n ) = 1 3 , so n = 6 . The smallest such N is N = 2 6 = 6 4 .
We expand to get that a b a + b = x 1 , so a b = x ( a + b ) . Thus a b − x ( a + b ) = 0 and a b − x ( a + b ) + x 2 = x 2 . Simplifying, we get ( a − x ) ( b − x ) = x 2 . Note that a − x and b − x must be positive integers, otherwise one of a or b would not be a positive integer. We first choose a − x as a positive factor of x 2 then b − x would follow. Thus f ( N ) is the number of ways to choose a factor of x 2 , which means that x 2 must have 13 positive integer factors.
Thus x 2 is a 12th power of a prime, thus its minimum is 2 1 2 = 4 0 9 6 . Hence the smallest possible value of N is 4 0 9 6 = 6 4 .
Let N = p k , where p is a prime and k is a positive integer. Then we have a 1 + b 1 = p k 1 . Clearing out the fractions, we get a p k + b p k = a b . Manipulating this equation allows us to obtain the factorization p 2 k = ( a − p k ) ( b − p k ) . Factoring the left-hand side as p 2 k = ( p i ) ( p 2 k − i ) for some integer i gives the solution ( a , b ) = ( p i + p k ) , ( p 2 k − i + p k ) . Because a and b are integers, we must have 0 ≤ i ≤ 2 k . In addition, we can't factorize p 2 k = ( − p i ) ( − p 2 k − i ) because then a and b would not both be positive; so, these are the only solutions.
For each integer i with 0 ≤ i ≤ 2 k , there is exactly one solution to the original equation. Thus, there are 2 k + 1 solutions to the original equation. Because we want f ( N ) = 1 3 , we solve the equation 2 k + 1 = 1 3 to get k = 6 . So, N is in the form p 6 for some prime p . To minimize N , we take p = 2 , hence the smallest possible value of N is 2 6 = 6 4 .
We can manipulate the equation:
a 1 + b 1 = x 1
a b a + b = x 1
x ( a + b ) = a b
a b − x ( a + b ) + x 2 = x 2
a b − a x − b x + x 2 = x 2
( a − x ) ( b − x ) = x 2
Since a , b > 0 , then a − x , b − x > − x . Suppose that a − x ≤ 0 , then b − x ≤ 0 (otherwise ( a − x ) ( b − x ) ≤ 0 while x 2 > 0 ). But then ( a − x ) ( b − x ) = ( x − a ) ( x − b ) < x ⋅ x = x 2 , impossible. So a − x > 0 . Similarly, b − x > 0 .
Now, a − x can be any positive factor of x 2 , and b − x will simply follow suit ( b − x = a − x x 2 ). Obviously as a − x is a factor of x 2 , a − x x 2 is a positive integer and hence a possible value for b − x . On the other hand, when a − x is not a factor of x 2 , a − x x 2 is not an integer and hence b − x will not be an integer, and hence b will not be an integer, contradiction. So there are precisely as many solutions of ( a , b ) as there are positive factors of x 2 . So f ( x ) is precisely τ ( x 2 ) where τ is the divisor function .
Now, let N = p n because N is a prime power. Then f ( N ) = f ( p n ) = τ ( p 2 n ) = 2 n + 1 , since there are 2 n + 1 divisors of p 2 n (namely p 0 , p 1 , … , p 2 n ). Since f ( N ) = 1 3 , we have 2 n + 1 = 1 3 and so n = 6 .
Since p 6 is strictly increasing for positive p , the smallest value of N is achieved when p is minimized. The smallest prime is 2 , so p = 2 gives the minimum value of N = 2 6 = 6 4 .
Remark: The problem originally states f ( N ) = 1 2 . After I complained (by the dispute feature, and there should be other people complaining too), the problem was changed to f ( N ) = 1 3 .
Log in to reply
I was stuck here for hours, no maybe days, and when the problem was suddenly corrected, I was ....................... (can't explain)
I complained as well. It took me an embarrassingly long time to realize that if (a,b) was a solution, then so was (b,a) so the solutions always occur in pairs, except for the case where a = b = 2x, which means that f(N) will always be an odd number.
We rewrite the equation:
a 1 + b 1 = x 1 ⟺ a b a + b = x 1 ⟺ a b = ( a + b ) x ⟺ x 2 = a b − ( a + b ) x + x 2 = ( a − x ) ( b − x )
Therefore, the pair ( a , b ) is a solution iff ( a − x ) ( b − x ) is a factorization of x 2 with a − x and b − x both larger than − x . This last condition rules out factorizations involving negative factors, since then one of them is necessarily at most − x . All factorizations with positive factors however yield a valid solution.
It is specified that x = p n . The factorizations of x 2 = p 2 n are all of the form p k p 2 n − k , with 0 ≤ k ≤ 2 n , hence there are 2 n + 1 factorizations and thus 2 n + 1 solutions to the original equation. Therefore, 2 n + 1 = 1 3 , and n = 6 .
The smallest number x of the form p 6 is trivially x = 2 6 = 6 4
As good as usual.Like your solution
N 1 = a 1 + b 1
So, N is equal to:
N= a + b a b
ab=aN + bN
ab - aN - bN=0
ab - aN - bN - N 2 =N 2
(a-N)(b-N)=N 2
N 2 is a perfect square and N is a prime power, therefore N 2 must have 13 divisors in order to f(N)=12, because each divisor correspond to one ordered pair(a,b), but N, because N 2 is a perfect square
Anyone that assumed that f(N) = 12 is wrong, because N^2 will always have a positive odd number of divisors. Therefore, f(N) must equal 13, if there was a mistake, and 64 is the wrong answer. There is in fact, no real answer to this question.
a 1 + b 1 = N 1 ⇒ ( a − N ) ( b − N ) = N 2 . Let N = p k . ( a − p k ) ( b − p k ) = p 2 k have clearly 2 k solutions. We want 1 2 solutions, so we set k = 6 . The minimum value of N is clearly 2 6 = 6 4 .
shouldn't f(p^k) = 2k+1?
a = b = 2N is always a solution for the equation, which makes the number of ordered pairs positive integer solutions always odd. So f(N) = 12 is impossible.
As mentioned in an earlier comment, f ( n ) = 1 3 , not 1 2 . f ( n ) is odd for all n . I asked for f ( n ) = 7 is my original question, and it has been changed to 1 2 here. Most possibly it is a typo.
Log in to reply
This is a modification of the 10th problem of INMO 1991. Right Sreejato?
Oops, miscounted. :(
Let N = p n . Thus we examine the equation a 1 + b 1 = p n 1 . Rearranging slightly, we have:
p n 1 = a b a + b
∴ p n = a + b a b
Let us say, without loss of generality, that a ≤ b . Since p n is an integer, a + b must divide a b . It follows that b must be a multiple of a and thus can be written as k a , for some integer k . Plugging in, we have:
a + b a b = a + k a k a 2 = 1 + k k a = p n
k is not a factor of 1 + k , thus 1 + k must divide a . Since p is prime, we have k = p c and 1 + k a = p n − c , for some integer c such that 0 ≤ c ≤ n . Solving for a and b , we have:
a = ( 1 + k ) p n − c = ( 1 + p c ) p n − c
b = k a = ( 1 + p c ) p n
Plugging in valid values for c ( c is an integer such that 0 ≤ c ≤ n ) will always result in a pair of integers ( a , b ) such that a 1 + b 1 = p n 1 . Additionally, for every ordered pair ( a , b ) obtained, we also have an additional pair ( b , a ) , except for when c = 0 , because then a = b . Since we want f ( N ) = 1 3 , we want c to take on 2 1 3 − 1 = 6 values, not including 0 . We thus we choose n = 6 .
The smallest prime power where n = 6 is 2 6 = 6 4 .
Rearranging the expression we got: b = a − x a x
So a − x ∣ a x , what implies that:
⇒ a − x ∣ a x − x ( a − x )
⇒ a − x ∣ x 2
Therefore, we have as many a 's satisfying the given conditions as divisors of x 2 . The smallest number that have exactly 1 3 divisors and is a prime power is 2 1 2 . Therefore, the number we're looking for is 2 1 2 = 2 6 = 6 4 .
Clearly, a , b > x . Multiply both sides of a 1 + b 1 = x 1 by a b x to get:
b x + a x = a b , i.e. 0 = a b − a x − b x .
Using SFFT we can add x 2 to both sides to get:
x 2 = a b − a x − b x + x 2 = ( a − x ) ( b − x ) .
Hence, a − x and b − x must be complementary positive factors of x 2 .
If x 2 has m positive factors, 1 = f 1 < f 2 < ⋯ < f m − 1 < f m = x 2 , then:
( a − x , b − x ) = ( f k , f m − k ) , i.e. ( a , b ) = ( x + f k , x + f m − k ) for k ∈ 1 , m .
There are m such pairs ( a , b ) . So, f ( x ) is the number of positive factors of x 2 .
If N = p n , then N 2 = p 2 n which has 2 n + 1 positive factors.
So, f ( N ) = f ( p n ) = 2 n + 1 = 1 3 ⇝ n = 6 . So, we need N = p 6 .
The smallest prime is p = 2 . Hence, the smallest such N is 2 6 = 6 4 .
We can try the prime power 2^6 or 64. (1/a) + (1/b) = 1/64. (a + b)/(ab) = 1/64. 64a + 64b = ab. a(b - 64) = 64b. From this equation we can find the a in term of b, namely: a = 64 + (64*64)/(b - 64). We can have 13 pairs of (a,b): b - 64 =1, 2,4,8,16,32,64,128,256,512,1024,2048,4096. So (a,b) = (4160,65),(2112,66),(1088,68),(576,72),(320,80),(192,96),(128,128),(96,192),(80,320),(72,576),(68,1088),(66,2112),(65,4160).
And to prove that 6 4 is the minimum value that works?
Rearranging gives:
- b = x + a − x x 2
We hence have positive solutions corresponding to positive factors of x^2. If x is a prime power p^n then f(x)=2n. We therefore want n=6 and 64 is the smallest sixth power.
N.B. We presumably we cannot have a=b because otherwise f(x) would need to be odd. Two orders for distinct solutions and just one order for a=2x, b=2x.
We have a 1 + b 1 = N 1 . Clearing denominators we get a b = N a + N b ⟹ a b − N a − N b = 0 ⟹ ( a − N ) ( b − N ) = N 2 . Note that the number of factors of N 2 is also going to be the number of ordered pairs (a,b) satisfying the equation. But wait! There are an even number of solutions, and an odd number of factors of N 2 ! Where did we go wrong? Notice that the ordered pair (0,0) satisfies our equation after clearing denominators, but it does not satisfy our original equation. We had assumed that a , b = 0 but included it in our count. Therefore, we must subtract this case from our total count. So we need to find an N such that the factors of N 2 minus one is equal to twelve. Therefore, the number of factors of N 2 is 13. We represent N = p n . So the number of factors of p 2 n is 13. Therefore 2 n + 1 = 1 3 and n = 6 . We want N as small as possible, so we make p = 2 . Therefore our final answer is 2 6 = 6 4
What went wrong in your argument is the contradiction: "There are an odd number of factors of N^2". Okay, I assume here you are only considering the positive factors. But then the pair (0,0) gives -N and -N for the factors of N^2. Therefore, your subtraction is invalid.
Log in to reply
I had the same problem, there is nothing in the question that states why you can't have a=b, so it's valid to.
Problem Loading...
Note Loading...
Set Loading...
Because all of a , b , x are positive, we can multiply through by a ⋅ b ⋅ x in the equation a 1 + b 1 = x 1 to get x ( a + b ) = a b . Rearranging and factoring gives us ( a − x ) ( b − x ) = x 2 , and from this, it is clear that f ( x ) is the number of divisors of x 2 , since ( a − x ) , ( b − x ) ∈ Z .
We have x = p n with p a positive prime and n a positive integer. The previous equation becomes ( a − p n ) ( b − p n ) = p 2 n . Clearly, p 2 n has 2 n + 1 divisors, and so f ( p n ) = 1 3 → 2 n + 1 = 1 3 → n = 6 . Thus, we want the smallest possible value of p 6 where p is a positive prime, which is obviously 2 6 = 6 4 .