Let P ( n ) denote the product of all positive submultiples of a natural number n .
There is a natural number N satisfying:
⎩ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎧ ⌊ lo g 2 N ⌋ = lo g 2 N ; 2 6 0 N is a natural number, but 2 8 5 N is not ; N ≡ 3 7 9 ( m o d 9 1 5 ) ; ⌊ lo g N P ( N ) ⌋ + 9 1 5 = 2 0 1 5 .
The smallest value for N is k = 1 ∏ n p k a k , such that for all natural numbers k ≤ n , ⎩ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎧ p k is a prime; p k − 1 < p k , where p 0 = 1 ; a k is a natural number.
Find the value of k = 1 ∑ n p k a k .
Notation: ⌊ ⋅ ⌋ denotes the floor function .
This problem was created by me, celebrating Undertale 's 2nd anniversary.
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.
A condensed version of my solution! It's really nice and organized :)
Define f ( n ) as the number of positive submultiples of a natural number n .
Lemma: P ( N ) = N 2 f ( N ) .
Let N = k = 1 ∏ m P k A k .
Let 8 4 = 2 2 × 3 × 7 .
.
Then let's enumerate all the submultiples of N , with all of them factorized.
2 0 × 3 0 × 7 0 , 2 0 × 3 1 × 7 0 , 2 0 × 3 0 × 7 1 , 2 0 × 3 1 × 7 1 , 2 1 × 3 0 × 7 0 , 2 1 × 3 1 × 7 0 , 2 1 × 3 0 × 7 1 , 2 1 × 3 1 × 7 1 , 2 2 × 3 0 × 7 0 , 2 2 × 3 1 × 7 0 , 2 2 × 3 0 × 7 1 , 2 2 × 3 1 × 7 1
.
For some arbitrary natural numbers i ≤ m and j ≤ A i , we will see that P i j is written exactly f ( P i A i N ) times.
2 0 × 3 0 × 7 0 , 2 0 × 3 1 × 7 0 , 2 0 × 3 0 × 7 1 , 2 0 × 3 1 × 7 1 , 2 1 × 3 0 × 7 0 , 2 1 × 3 1 × 7 0 , 2 1 × 3 0 × 7 1 , 2 1 × 3 1 × 7 1 , 2 2 × 3 0 × 7 0 , 2 2 × 3 1 × 7 0 , 2 2 × 3 0 × 7 1 , 2 2 × 3 1 × 7 1
.
Which means, if we multiply all of them, P i j will be multiplied f ( P i A i N ) = A i + 1 k = 1 ∏ n ( A k + 1 ) times.
Think N as a factorized form. P i A i N contains everything except P i A i , and therefore, all ( A k + 1 ) ’s are going to be multiplied except ( A i + 1 ) .
.
So if we multiply all P i j 's, the product would be ( P i j ) A i + 1 f ( N ) = ( P i A i + 1 f ( N ) ) j .
Remember, k = 1 ∏ n ( A k + 1 ) = f ( N ) .
.
Since this applies for all j ≤ A i , if we multiply all the factors with base P i , the product would be ( P i A i + 1 f ( N ) ) 2 A i ( A i + 1 ) = P i 2 f ( N ) ⋅ A i .
0 + 1 + 2 + ⋯ + j + ( j + 1 ) + ⋯ + A i = 2 A i ( A i + 1 ) .
.
Since this applies for all i ≤ m , if we multiply all the submultiples of N , (!!!) the product would be P ( N ) = i = 1 ∏ m P i 2 f ( N ) ⋅ A i = i = 1 ∏ m ( P i A i ) 2 f ( N ) = N 2 f ( N ) .
Therefore the lemma is proven.
Now we see that ⌊ lo g N P ( N ) ⌋ = ⌊ 2 f ( N ) ⌋ = 2 0 1 5 − 9 1 5 = 1 1 0 0 .
However, if f ( N ) = 2 2 0 0 , we need 2 k ( 6 0 ≤ k ≤ 8 4 ) multiplied inside N however, 2 2 0 0 doesn't have a submultiple that is between 6 1 and 8 5 .
CONDITION: 2 6 0 N is a natural number, but 2 8 5 N is not.
.
So we need f ( N ) = 2 2 0 1 = 3 1 × 7 1 , so that 2 7 0 is multiplied inside N .
Values of N would be, then, 2 7 0 × 7 3 0 , 2 7 0 × 1 1 3 0 , 2 7 0 × 1 3 3 0 , . . .
CONDITION: N ≡ 3 7 9 ( m o d 9 1 5 ) ; which implies that N is not a multiple of 3 or 5.
.
Note that N ≡ 1 ( m o d 3 ) , N ≡ 4 ( m o d 5 ) , N ≡ 1 3 ( m o d 6 1 ) .
9 1 5 = 3 × 5 × 6 1 ; 3 7 9 ≡ 1 ( m o d 3 ) , 3 7 9 ≡ 4 ( m o d 5 ) , 3 7 9 ≡ 1 3 ( m o d 6 1 )
Let N = 2 7 0 × 7 3 0 . Then,
N ≡ 4 × 9 ≡ 6 ( m o d 1 0 ) ; N ≡ 1 ( m o d 5 ) .
This is inconsistent with the condition, so this is not the correct answer.
.
Let N = 2 7 0 × 1 1 3 0 . Then,
N ≡ 1 ( m o d 3 ) . square of a natural number is always equivalent to 1 in mod 3. N ≡ 4 × 1 ≡ 4 ( m o d 1 0 ) ; N ≡ 4 ( m o d 5 ) . N ≡ 2 1 0 × 1 2 1 1 5 ≡ 1 2 8 × 8 × ( − 1 ) 1 5 ≡ 6 × 8 × ( − 1 ) ≡ − 4 8 ≡ 1 3 ( m o d 6 1 ) . ϕ ( 6 1 ) = 6 0 . 1 2 1 ≡ − 1 ( m o d 6 1 ) . 1 2 8 ≡ 6 ( m o d 6 1 ) .
This is consistent with all the conditions given, and thus, the minimum value for N is 2 7 0 × 1 1 3 0 .
∴ k = 1 ∑ n p k a k = 2 × 7 0 + 1 1 × 3 0 = 4 7 0 .
You can also prove the lemma like this If a is a positive submultiple of N, then N/a is also a submultiple of N since N/a is an integer and N/(N/a)=a is also an integer. So letting a be each of f(N) possible submultiples of N, we get f(N) values of N/a which also are the submultiples of N. These f(N) values of N/a are exactly all the submultiples of N and therefore multiplying all these also gives P(N). So P(N) P(N)=((a) (N/a))^(f(N))=N^(f(N)) => P(N)=N^(f(N)/2)
I had to look up the term "submultiple", which I'd never heard of. In future, I'd use the term "divisor" or "factor", which is much more common.
Brilliant problem and solution.
Problem Loading...
Note Loading...
Set Loading...
If m , n are coprime positive integers, then a positive integer a divides m n if and only if a = b c where b , c are positive integers where b divides m and c divides n . Thus P ( m n ) = b ∣ m , c ∣ n ∏ b c = b ∣ m ∏ b σ 0 ( n ) P ( n ) = P ( m ) σ 0 ( n ) P ( n ) σ 0 ( m ) where σ 0 ( n ) is the number of divisors of n . This means that Q ( n ) = P ( n ) σ 0 ( n ) 1 is a multiplicative function. Since Q ( p a ) = ( j = 0 ∏ a p j ) a + 1 1 = p 2 1 a = p a for all primes p and positive integers a , we deduce that Q ( n ) = n for all n ≥ 1 , and hence P ( n ) = n 2 1 σ 0 ( n ) . Note that σ 0 ( n ) is even except when n is a perfect square, and so this formula for P ( n ) always yields an integer, thankfully!
The last condition tells us that ⌊ 2 1 σ 0 ( N ) ⌋ = 1 1 0 0 , and hence that σ 0 ( N ) = 2 2 0 0 or 2 2 0 1 . The second condition tells us that the index of 2 in N lies between 6 0 and 8 4 , and hence σ 0 ( N ) is divisible by some integer between 6 1 and 8 5 . There is no such factor of 2 2 0 0 , and hence σ 0 ( N ) = 2 2 0 1 = 3 1 × 7 1 . Thus it follows that N = 2 7 0 × p 3 0 for some odd prime p . This makes the first condition automatically true.
The smallest prime p for which 2 7 0 × p 3 0 ≡ 3 7 9 ( m o d 9 1 5 ) is p = 1 1 , and hence the answer is 2 × 7 0 + 1 1 × 3 0 = 4 7 0 .