Let s ( n ) be the sum of the digits of n 's base- 2 0 1 7 representation. Compute the sum n = 1 ∑ ∞ n ( n + 1 ) s ( n ) .
Round your answer to the nearest thousandth.
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.
That's some significant math. I didn't follow all of it, but I was impressed.
I ended up using a Python program, but it was taking a looooonnnnnggg time to converge.
Let s p ( n ) be the sum of the digits of n in the base p representation of n . Note that s p ( n ) = O ( ln n ) as m → ∞ , and hence it is clear that the infinite serie S p = n ≥ 1 ∑ n ( n + 1 ) s p ( n ) converges.
Let j p ( n ) be the index of p in n . It is easy to see that s p ( n ) − s p ( n − 1 ) = 1 − ( p − 1 ) j p ( n ) n ≥ 1 and so, since n = 1 ∑ ∞ n ( n + 1 ) s p ( n ) = n ≥ 1 ∑ n s ( n ) − s ( n − 1 ) = n ≥ 1 ∑ n 1 − ( p − 1 ) j p ( n ) we see that n − 1 gets added to the sum once for each n , but ( p − 1 ) n − 1 is subtracted from the sum once each time p divides n . Thus n = 1 ∑ p K n ( n + 1 ) s p ( n ) = n = 1 ∑ p K n 1 − ( p − 1 ) n = 1 ∑ p K − 1 p n 1 − ( p − 1 ) n = 1 ∑ p K − 2 p 2 n 1 − ⋯ − ( p − 1 ) n = 1 ∑ 1 p K n 1 = H p K − ( p − 1 ) j = 1 ∑ K p j H p K − j Now we use the fact that X n = H n − ln n − γ = O ( n − 1 ) n → ∞ and calculate ln ( p K ) − ( p − 1 ) j = 1 ∑ K p j ln ( p K − j ) 1 − ( p − 1 ) j = 1 ∑ K p − j = p − 1 p ( 1 − p − K ) ln p = p − K so that n = 1 ∑ p K n ( n + 1 ) s p ( n ) = p − 1 p ( 1 − p − K ) ln p + γ p − K + [ X p K − ( p − 1 ) j = 1 ∑ K p j X p K − j ] Since ∣ X n ∣ ≤ A n − 1 for some constant A , we deduce that ∣ ∣ ∣ ∣ ∣ ∣ n = 1 ∑ p K n ( n + 1 ) s p ( n ) − p − 1 p ln p ∣ ∣ ∣ ∣ ∣ ∣ ≤ [ p − 1 p ln p + γ + A ] p − K + ( p − 1 ) j = 1 ∑ K p K − j × p j A ≤ [ p − 1 p ln p + γ + A + A K ( p − 1 ) ] p − K for all integers K . Thus we deduce that S p = p − 1 p ln p . With p = 2 0 1 7 , we obtain the answer 7 . 6 1 3 1 4 1 0 2 5 .
I couldn't decide which solution I liked better, so I upvoted them both. This one is pretty much exactly what I did, except it's far more scrupulous about convergence issues; the OP's solution seems to avoid those issues by the nice trick of replacing n by n s and later taking the limit as s → 1 + .
Problem Loading...
Note Loading...
Set Loading...
Let p = 2 0 1 7 . We will use the following lemmas:
Lemma 1 v p ( n ! ) = p − 1 n − s ( n ) Lemma 2 For s > 1 , s − 1 s > ζ ( s ) > s − 1 1 The lemma follows from comparing ζ ( s ) to the area under the curve f ( t ) = t − s , giving us 1 + ∫ 1 ∞ t − s d t > ζ ( s ) > ∫ 1 ∞ t − s . Corollary lim s → 1 + ( s − 1 ) ζ ( s ) = 1 .
We come back to the main problem. Observe that n ≥ 1 ∑ n ( n + 1 ) s ( n ) = n ≥ 1 ∑ n s ( n ) − n + 1 s ( n ) = n = 1 ∑ ∞ n s ( n ) − s ( n − 1 ) = n = 1 ∑ ∞ n 1 − ( p − 1 ) v p ( n ) .
We define the function F ( s ) = n = 1 ∑ ∞ n s 1 − ( p − 1 ) v p ( n ) for s > 1 . This simplifies to F ( s ) = ζ ( s ) − ( p − 1 ) n = 1 ∑ ∞ n s v p ( n ) = ζ ( s ) − ( p − 1 ) k = 1 ∑ ∞ m = 1 ∑ ∞ ( p k m ) s 1 = ζ ( s ) − ( p − 1 ) ζ ( s ) ( 1 − p − s p − s ) = ζ ( s ) ( p s − 1 p s − p ) . Now, we take the limit as s → 1 + , s → 1 + lim F ( s ) = p − 1 1 s → 1 + lim ζ ( s ) ( s − 1 ) s − 1 p s − p = p − 1 p s → 1 + lim s − 1 p s − 1 − 1 = p − 1 p lo g p where the last equality follows from L'Hospital's Rule.