Inspired by Priyanshu Mishra 2

Consider any sequence of n ( 1 ) n \,(\geq 1) positive integers { a 1 , a 2 , , a n } \{a_1, a_2, …, a_{n}\} , and extend it periodically to an infinite sequence { a 1 , a 2 , } \{a_1, a_2, …\} by defining a i + n = a i a_{i+n} = a_i for all i 1 i \geq 1 . The sequence { a 1 , a 2 , , a n } \{a_1, a_2, …, a_{n}\} is called n n -Mishra-interesting if

  • 1 a 1 a 2 a n a 1 + n 1 \leq a_1 \leq a_2 \leq \cdots \leq a_{n} \leq a_1 + n , and
  • a a i i a_{a_i} \leq i for i = 1 , 2 , . . . , n i=1,2,...,n .

Let N n N_n be the number of n n -Mishra-interesting sequences { a 1 , a 2 , , a n } \{a_1, a_2, …, a_{n}\} . Then lim n N n 4 n n 3 / 2 = a b π 1 / c \underset{n\to \infty }{\text{lim}} N_n 4^{-n} n^{3/2}=\frac{a}{b \pi ^{1/c}} for positive integers a , b , c a,b,c satisfying gcd ( a , b ) = 1 \text{gcd}(a,b) = 1 .

Submit a + b + c a+b+c .



The answer is 9.

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.

1 solution

Stephen Phillips
Jan 30, 2019

The full explanation is way too long so I will give the summary here (proofs omitted as they are tedius):

Lemma 1: The following is equivalent to n-Mishra Interesting

  • (1) a 1 = 1 a_1 = 1
  • (2) a i 1 a i i or a i = n + 1 , i { 2 , , n } a_{i-1} \leq a_i \leq i \textrm{ or } a_i = n + 1, \; \forall i \in \{ 2, \ldots, n\}

This is a much easier definition to work with, as we can actually count these. First consider the sequences that have no entries that are n + 1 n + 1 . These correspond exactly to the Catalan numbers, specifically the interpretation of lattice paths below the diagonal, with the values of the sequence being the height of the columns. See here on Wikipedia to learn more about them. That means the number of n-Mishra interesting sequences of length n with no entries that have value n + 1 n+1 is exactly 1 n + 1 ( 2 n n ) \frac{1}{n+1} \binom{2n}{n} Now note that if a j = n + 1 a_j = n + 1 then all subsequent entries in the sequence are n + 1 n+1 . Thus we count all of them that have a 2 = n + 1 a_2 = n + 1 , then a 3 = n + 1 a_3 = n + 1 (and a 2 n + 1 a_2 \neq n + 1 ), etc. all the way to only a n = n + 1 a_n = n + 1 . This is exactly the same as the cases with no values of n + 1 n + 1 but for shorter sequences (lengths 1, 2, ... n-1). So to count all of them, we sum them all of them (and the number without n + 1 n + 1 in the sequence) to get our final count: N n = k = 1 n 1 k + 1 ( 2 k k ) N_n = \sum_{k=1}^{n} \frac{1}{k+1} \binom{2k}{k} .

Now we need to use Stirling's approximation to simplify 1 k + 1 ( 2 k k ) 1 k + 1 4 k π k \frac{1}{k+1} \binom{2k}{k} \approx \frac{1}{k+1} \frac{4^k}{\sqrt{\pi k}} , which is true as k k becomes large. So we get: lim n N n 4 n n 3 / 2 = lim n ( k = 1 n 1 k + 1 ( 2 k k ) ) 4 n n 3 / 2 = lim n ( k = 1 n 1 k + 1 4 k π k ) 4 n n 3 / 2 \lim_{n \rightarrow \infty} N_n 4^{-n} n^{3/2} = \lim_{n \rightarrow \infty} \left( \sum_{k=1}^{n} \frac{1}{k+1} \binom{2k}{k} \right) 4^{-n} n^{3/2} = \lim_{n \rightarrow \infty} \left( \sum_{k=1}^{n} \frac{1}{k+1} \frac{4^k}{\sqrt{\pi k}} \right) 4^{-n} n^{3/2}

= 1 π 1 / 2 lim n k = 1 n n 3 / 2 k 3 / 2 + k 1 / 2 1 4 n k = 1 π 1 / 2 lim n k = 0 n 1 n 3 / 2 ( n k ) 3 / 2 + ( n k ) 1 / 2 1 4 k = \frac{1}{\pi^{1/2}} \lim_{n \rightarrow \infty} \sum_{k=1}^{n} \frac{n^{3/2}}{k^{3/2}+k^{1/2}} \frac{1}{4^{n-k}} = \frac{1}{\pi^{1/2}} \lim_{n \rightarrow \infty} \sum_{k=0}^{n-1} \frac{n^{3/2}}{(n-k)^{3/2}+(n-k)^{1/2}} \frac{1}{4^k}

To evaluate that last sum, you can prove that as the fractional term in front of the 1 / 4 k 1 / 4^{k} can be ignored ( this is intuitive I actually had a hard time doing this rigorously). So that simplifies to: 1 π 1 / 2 lim n k = 0 n 1 1 4 k = 4 π 1 / 2 3 \frac{1}{\pi^{1/2}} \lim_{n \rightarrow \infty} \sum_{k=0}^{n-1} \frac{1}{4^k} = \frac{4}{\pi^{1/2} 3} . Thus we have a = 4 a = 4 , b = 3 b = 3 , and c = 2 c = 2 , thus a + b + c = 9 a + b + c = 9 , our final answer.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...