Interesting Limit (8)

Calculus Level 3

lim n k = 0 n ln ( n k ) n 2 = ? \large\lim_{n\to\infty}\frac{\displaystyle \sum_{k=0}^n\ln\binom nk}{n^2}=\, ?


The answer is 0.5.

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.

2 solutions

Leonel Castillo
Aug 15, 2018

k = 0 n log ( n k ) = log k = 0 n ( n k ) = log ( n ! ) n + 1 ( 2 ! × . . . × n ! ) 2 = log ( 2 3 n × 3 5 n × . . . × n 2 n 1 n ) = k = 2 n ( 2 k 1 n ) log k \sum_{k=0}^n \log \binom{n}{k} = \log \prod_{k=0}^n \binom{n}{k} = \log \frac{(n!)^{n+1}}{(2! \times ... \times n!)^2} = \log \left( 2^{3-n} \times 3^{5-n} \times ... \times n ^{2n - 1 - n} \right) = \sum_{k=2}^{n} (2k - 1 - n) \log k

Diving by n 2 n^2 we get three series: k = 2 n 2 k log k n 2 k = 2 n log k n 2 k = 2 n log k n \sum_{k=2}^n \frac{2k \log k}{n^2} - \sum_{k=2}^n \frac{\log k}{n^2} - \sum_{k=2}^n \frac{\log k}{n} First, notice that k = 2 n log k n 2 = log ( n ! ) n 2 log ( n n ) n 2 = log n n 0 \sum_{k=2}^n \frac{\log k}{n^2} = \frac{\log(n!)}{n^2} \leq \frac{\log(n^n)}{n^2} = \frac{\log n}{n} \to 0 so the middle term vanishes. Now, remembering Euler's summation formula in the form k = 2 n f ( k ) = 2 n f ( x ) d x + O ( f ( n ) ) \sum_{k=2}^n f(k) = \int_2^n f(x) dx + O(f(n)) , if we apply this to the functions f ( x ) = 2 x log x f(x) = 2x \log x and f ( x ) = log ( x ) f(x) = \log(x) we notice that the error terms f ( n ) n 2 \frac{f(n)}{n^2} vanish in both cases. Thus we can say that in the limit: lim n k = 2 n 2 k log k n 2 k = 2 n log k n = lim n 1 n 2 2 n 2 x log x d x 1 n 2 n log x d x = lim n 1 2 + log n + 2 log 4 n 2 log n + 1 2 log 4 n = 1 2 + 1 = 1 2 \lim_{n \to \infty} \sum_{k=2}^n \frac{2k \log k}{n^2} - \sum_{k=2}^n \frac{\log k}{n} = \lim_{n \to \infty} \frac{1}{n^2} \int_2^n 2x \log x dx - \frac{1}{n} \int_2^n \log x dx = \lim_{n \to \infty} -\frac{1}{2} + \log n + \frac{2 - \log 4}{n^2} - \log n + 1 - \frac{2 - \log 4}{n} = -\frac{1}{2} + 1 = \frac{1}{2}

Note: What has actually been proven is the asymptotic formula k = 0 n log ( n k ) = n 2 2 + O ( n log n ) \sum_{k=0}^n \log \binom{n}{k} = \frac{n^2}{2} + O(n \log n)

Brian Lie
Apr 21, 2018

Relevant wiki: Stolz–Cesàro theorem

L = lim n k = 0 n ln ( n k ) n 2 Using Stolz-Ces a ˋ ro theorem = lim n k = 0 n + 1 ln ( n + 1 k ) k = 0 n ln ( n k ) ( n + 1 ) 2 n 2 = lim n k = 0 n ln ( n + 1 k ) ( n k ) + ln ( n + 1 n + 1 ) 2 n + 1 = lim n k = 0 n ln n + 1 n k + 1 2 n + 1 = lim n ( n + 1 ) ln ( n + 1 ) k = 1 n + 1 ln k 2 n + 1 Using Stolz-Ces a ˋ ro theorem again = lim n ( n + 1 ) ln ( n + 1 ) n ln n ln ( n + 1 ) ( 2 n + 1 ) ( 2 n 1 ) = lim n ln ( ( 1 + 1 n ) n ) 2 = ln e 2 = 0.5 \begin{aligned} L&=\lim_{n\to\infty}\frac{\displaystyle\sum_{k=0}^n\ln\binom nk}{n^2}&\small\color{#3D99F6}\text{Using Stolz-Cesàro theorem} \\&=\lim_{n\to\infty}\frac{\displaystyle\sum_{k=0}^{n+1}\ln\binom {n+1}k-\sum_{k=0}^n\ln\binom nk}{(n+1)^2-n^2} \\&=\lim_{n\to\infty}\frac{\displaystyle\sum_{k=0}^n\ln\dfrac{\binom {n+1}k}{\binom nk}+\ln\binom {n+1}{n+1}}{2n+1} \\&=\lim_{n\to\infty}\frac{\displaystyle\sum_{k=0}^n\ln\frac{n+1}{n-k+1}}{2n+1} \\&=\lim_{n\to\infty}\frac{(n+1)\ln (n+1)-\displaystyle\sum_{k=1}^{n+1}\ln k}{2n+1}&\small\color{#3D99F6}\text{Using Stolz-Cesàro theorem again} \\&=\lim_{n\to\infty}\frac{(n+1)\ln (n+1)-n\ln n-\ln (n+1)}{(2n+1)-(2n-1)} \\&=\lim_{n\to\infty}\frac{\ln\left(\left(1+\dfrac 1n\right)^n\right)}2 \\&=\frac{\ln e}2=\boxed{0.5} \end{aligned}

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...