Evaluate $$\lim_{n\to\infty} \left(\frac{1^p+2^p+3^p + \cdots + n^p}{n^p} - \frac{n}{p+1}\right)$$
-
2This is Problem 2.3.12 (g) from Kaczor-Nowak: Problems in Mathematical Analysis I. The problem is stated on p.37 and a solution is given on p.184. This solution uses Stolz-Cesaro Theorem. – Martin Sleziak May 25 '12 at 10:36
7 Answers
The result is more general.
Fact: For any function $f$ regular enough on $[0,1]$, introduce $$ A_n=\sum_{k=1}^nf\left(\frac{k}n\right)\qquad B=\int_0^1f(x)\mathrm dx\qquad C=f(1)-f(0) $$ Then, $$ \lim\limits_{n\to\infty}A_n-nB=\frac12C $$
For any real number $p\gt0$, if $f(x)=x^p$, one sees that $B=\frac1{p+1}$ and $C=1$, which is the result in the question.
To prove the fact stated above, start from Taylor formula: for every $0\leqslant x\leqslant 1/n$ and $1\leqslant k\leqslant n$, $$ f(x+(k-1)/n)=f(k/n)-(1-x)f'(k/n)+u_{n,k}(x)/n $$ where $u_{n,k}(x)\to0$ when $n\to\infty$, uniformly on $k$ and $x$, say $|u_{n,k}(x)|\leqslant v_n$ with $v_n\to0$. Integrating this on $[0,1/n]$ and summing from $k=1$ to $k=n$, one gets $$ \int_0^1f(x)\mathrm dx=\frac1n\sum_{k=1}^nf\left(\frac{k}n\right)-\frac1n\int_0^{1/n}u\mathrm du\cdot\sum_{k=1}^nf'\left(\frac{k}n\right)+\frac1nu_n $$ where $|u_n|\leqslant v_n$. Reordering, this says that $$ A_n=nB+\frac12\frac1n\sum_{k=1}^nf'\left(\frac{k}n\right)-u_n=nB+\frac12\int_0^1f'(x)\mathrm dx+r_n-u_n $$ with $r_n\to0$, thanks to the Riemann integrability of the function $f'$ on $[0,1]$. The proof is complete since $r_n-u_n\to0$ and the last integral is $f(1)-f(0)=C$.

- 279,727
-
2+1. The all powerful Euler–Maclaurin!(http://en.wikipedia.org/wiki/Euler%E2%80%93Maclaurin_formula) – May 24 '12 at 11:47
This is a nice little question. I am assuming that $p \in \mathbb{Z}^+$, though same could be said about it when $p \notin \mathbb{Z}^+$. Before getting to the answer lets experiment a bit for small positive integers $p$. To start off, you could try for some values $p$.
For $p=1$, we get $$\lim_{n \rightarrow \infty} \left(\frac{ \dfrac{n(n+1)}{2}}{n} - \frac{n}{1+1} \right) = \lim_{n \rightarrow \infty} \frac12 = \frac12$$
For $p=2$, we get $$\lim_{n \rightarrow \infty} \left(\frac{ \dfrac{n(n+1)(2n+1)}{6}}{n^2} - \frac{n}{2+1} \right) = \lim_{n \rightarrow \infty} \left(\frac{(n+1)(n+1/2)}{3n} - \frac{n}3 \right)\\ = \lim_{n \rightarrow \infty} \left(\frac{n}3 + \frac12 + \frac1{6n} - \frac{n}3 \right)= \frac12$$
For $p=3$, we get $$\lim_{n \rightarrow \infty} \left(\frac{ \dfrac{n^2(n+1)^2}{4}}{n^3} - \frac{n}{3+1} \right) = \lim_{n \rightarrow \infty} \left(\frac{n^2 + 2n + 1}{4n} - \frac{n}4 \right)\\ = \lim_{n \rightarrow \infty} \left(\frac{n}4 + \frac12 + \frac1{4n} - \frac{n}4 \right)= \frac12$$
Hence, we would guess that it is $\dfrac12$ independent of $p$. And this turns out to be right.
Let us denote $1^p + 2^p + \cdots n^p = P_p(n)$. This is a polynomial of degree $p+1$ and is given by $$P_p(n) = \frac1{p+1} \sum_{k=0}^p \dbinom{p+1}{k} B_k n^{p+1-k}$$ where $B_k$ are the Bernoulli numbers. These polynomials are related to the Bernoulli polynomials and there are some really nice results on these polynomials and more can be found here.
Hence, $$\dfrac{P_p(n)}{n^{p}} = \dfrac1{p+1} \sum_{k=0}^p \dbinom{p+1}{k} B_k n^{1-k} = \dfrac1{p+1} \left(B_0 n + (p+1) B_1 + \mathcal{O} \left(\frac1n\right) \right)$$ where $B_0 = 1$ and $B_1 = \frac12$. What you are looking for is $$\lim_{n \rightarrow \infty} \left(\dfrac{P_p(n)}{n^{p}} - \dfrac{n}{p+1} \right) = \lim_{n \rightarrow \infty} \left(\dfrac1{p+1} \left(n + (p+1) B_1 + \mathcal{O} \left(\frac1n\right) \right) - \dfrac{n}{p+1} \right)\\ = \lim_{n \rightarrow \infty} \left(B_1 + \mathcal{O} \left(\dfrac1n \right)\right)= B_1 = \frac12$$ independent of $p$.
Users Did and Ragib Zaman have provided excellent solutions. You might also want to look at Euler–Maclaurin formula which is of significance in this context.

- 279,727
-
1I very much doubt the above very nice proof is appropriate for high school level. I don't think there are many education systems around the globe that include big "o" notation, Bernoulli numbers and polynomials in their H.S. curriculum... – DonAntonio May 24 '12 at 11:38
-
@DonAntonio The big O notation was only used as a short hand. It could have as well been written out. Also, Bernoulli numbers and polynomials are taught in high school. – May 24 '12 at 11:51
-
I agree about the big O notation. About Bernoulli numbers and pol's: I thoroughly know the education system in Israel and Mexico and, to a less extent, in USA. None of them includes anything even close to Bernoulli pol's/numbers in general (which means: perhaps some elite, most probably private high schools, do. I know none but I can't discard them). I also know a little about the system in Spain and it's the same as far as I know. – DonAntonio May 24 '12 at 11:56
-
Kudos for the indian h.s. education. In my globe's zone this kind of stuff is not even mentioned, let alone worked on. – DonAntonio May 24 '12 at 12:05
-
@Chris Kindly accept one of the answers if you have understood and obtained what you want. You may want to read here (http://meta.stackexchange.com/questions/5234/how-does-accepting-an-answer-work) on how and why to accept answers. Try to accept answers for the other questions you have asked on the site as well, if you are satisfied with the quality of answers. – May 24 '12 at 16:40
-
@Chris Often there are lot of good answers and it gets difficult for the questioner to choose which one to accept :). But do accept one, since it is over all a good thing for the entire community. – May 24 '12 at 17:05
-
@Marvis, nice answer. One nitpick: it might be more clear to write $(p+1)B_1$ instead of $B_1(p+1)$ to avoid confusion between the Bernoulli numbers and the Bernoulli polynomials you mentioned. – Antonio Vargas May 25 '12 at 15:03
-
-
2
-
Well... its my answer. I am copying it onto my local machine. Why should stack exchange benefit out of my answers? – Dec 09 '13 at 06:20
-
@user17762 Why not? They clearly help people, as evidenced by the upvotes. Deleting them seems purely spiteful. – Potato Dec 09 '13 at 06:24
-
1@Potato If people (both the users and those who run stack-exchange) are not nice, then why should I waste my time in answering questions and providing answers? – Dec 09 '13 at 06:25
-
1@user17762 I'm sorry people have not been nice. That is unfortunate. But why deprive others, many of them quite nice people, of your answers? – Potato Dec 09 '13 at 06:26
-
If we draw the graph of $x^p$ from $x=1$ to $x=n,$ divide it into unit length intervals and approximate each segment of area by a trapezium (this is known as the trapezoidal rule) then we see that $$\int^n_1 x^p dx \approx \sum_{k=1}^n k^p - \frac{n^p+1}{2}.$$ The integral on the left is precisely $\displaystyle \frac{n^{p+1} -1}{p+1},$ so for large $n$ (where the major contribution is from the dominant terms) we have $$\sum_{k=1}^n k^p \approx \frac{n^{p+1}}{p+1} + \frac{n^p}{2}$$ so your limit is $1/2.$
For a precise solution, we need the error term along with the trapezoidal rule, which is derived here. It gives : $$\int^b_a f(x) dx = \frac{b-a}{2} ( f(a) + f(b) ) - \frac{(b-a)^3 }{12} f''(\zeta) $$ for some $\zeta \in [a,b].$ For $f(x)=x^p$ we have $f''(x) = p (p-1)x^{p-2}$ which is largest at $x=b$, the right end point. So the sum of the error terms in our application of the trapezoidal rule is at largest $$\frac{p(p-1)}{12} (2^{p-2} + 3^{p-2} + \cdots + n^{p-2}).$$ The sum in the brackets is overestimated by $\int^{n+1}_1 x^{p-2} dx= \frac{(n+1)^{p-1}-1}{p-1},$ so we get that $$\sum_{k=1}^n k^p = \frac{n^{p+1}}{p+1} + \frac{n^p}{2} + E_n$$ where $E_n$ is an error term that satisfies $\displaystyle \lim_{n\to\infty} \frac{E_n}{n^p} = 0$ which proves your limit.

- 35,127
-
2+1 Now this looks as something an advanced student in a very good (mathematicwise) high school could grasp. – DonAntonio May 24 '12 at 12:00
-
@Chris I just looked over my post and realized a horrible slip of the mind, it's no wonder why you didn't understand it before! The only wonder is how it got 7 upvotes in that condition lol. It's fine now though. Check out the edit and edit history for clarification. – Ragib Zaman May 25 '12 at 00:56
-
@RagibZaman Only the ideas are important. Yes you made a small algebraic error before but that doesn't really matter. :) – May 25 '12 at 03:50
-
I thought of $\int^n_1 x^p dx \approx \sum_{k=1}^n k^p - \frac{n^p+1}{2}$ but this doesn't modify the final result. – user 1591719 May 25 '12 at 10:17
-
-
1@Ragib Zaman: i totally agree with what Marvis said: "Only the ideas are important". It's OK. – user 1591719 May 25 '12 at 10:21
-
You have to be more careful with your limits! What does $f(n) \approx g(n)$ mean? If you want it to mean the same as $f(n)-g(n) \approx 0$, then your statement of the trapezoidal rule is false. If it doesn't mean the same as $f(n)-g(n) \approx 0$, then your conclusion is false. – TonyK May 25 '12 at 10:27
-
@Chris: It's really up to Ragib to provide more details. What does the $\approx$ sign in the first equation mean, precisely? – TonyK May 25 '12 at 11:07
-
I thought it was understood my estimates and overall solution was not precise, merely something to satisfy a high schooler. I used $\approx$ to denote dominant behaviour in the way some might say that for large $n$, $n^{10} + n \approx n^{10}$ even though the difference goes to infinity. It doesn't matter if the error goes to infinity, as long as it does so more slowly than $n^p$ so that after dividing by $n^p$ and taking limits, that term is still $0.$ – Ragib Zaman May 25 '12 at 12:44
-
If you want to give a water tight solution that is still understandable to a high schooler, you will have to spend a bit more time to derive the error estimate that goes along with the trapezoidal rule. Use integration by parts a few times to show them how Taylor series arise, then show them this. It shows the error in using the trapezoidal rule on $f$ is (up to constant factors) is $f''$, so for our application, the error is at most $n^{p-2}$, which suffices to find this limit. – Ragib Zaman May 25 '12 at 12:48
-
@Chris: Just so you understand: this is not a proof! It's just a motivation, which suggests what the correct answer might be. In particular, Ragib's "up to constant factor" (in the last comment) is wrong. The factor depends on the end-points of the integration range, so it depends on $n$. – TonyK May 25 '12 at 13:25
-
-
another method, using Stolz–Cesàro theorem: let ${ x }_{ n }=\left( p+1 \right) \left( { 1 }^{ p }+{ 2 }^{ p }+...+{ n }^{ p } \right) -{ n }^{ p+1 },{ y }_{ n }=\left( p+1 \right) { n }^{ p }$ $$\lim _{ x\rightarrow \infty }{ \frac { { x }_{ n+1 }-{ x }_{ n } }{ { y }_{ n+1 }-{ y }_{ n } } = } \lim _{ x\rightarrow \infty }{ \frac { \left( p+1 \right) { \left( n+1 \right) }^{ p }-{ \left( n+1 \right) }^{ p+1 }+{ n }^{ p+1 } }{ \left( p+1 \right) \left( { \left( n+1 \right) }^{ p }-{ n }^{ p } \right) } = } \\ =\lim _{ x\rightarrow \infty }{ \left( \frac { \left( p+1 \right) \left( { n }^{ p }+p{ n }^{ p-1 }+\frac { p\left( p-1 \right) }{ 2 } { n }^{ p-2 }+...+1 \right) }{ \left( p+1 \right) \left( { n }^{ p }+p{ n }^{ p-1 }+\frac { p\left( p-1 \right) }{ 2 } { n }^{ p-2 }+...+1-{ n }^{ p } \right) } \right) + } \\ +\frac { -{ n }^{ p+1 }-\left( p+1 \right) { n }^{ p }-\frac { p\left( p+1 \right) }{ 2 } { n }^{ p-1 }-...-1+{ n }^{ p+1 } }{ \left( p+1 \right) \left( { n }^{ p }+p{ n }^{ p-1 }+\frac { p\left( p-1 \right) }{ 2 } { n }^{ p-2 }+...+1-{ n }^{ p } \right) } $$ let's cobmine all coefficients of n,then divide numerator and denominator by $n^{ p-1 }$ and define sum of the all terms no more -1 power with $o\left( \frac { 1 }{ n } \right) $ $$\\ \\ \lim _{ x\rightarrow \infty }{ \frac { { x }_{ n+1 }-{ x }_{ n } }{ { y }_{ n+1 }-{ y }_{ n } } = } \lim _{ x\rightarrow \infty }{ \frac { \frac { p\left( p+1 \right) }{ 2 } +o\left( \frac { 1 }{ n } \right) }{ p\left( p+1 \right) +\left( \frac { 1 }{ n } \right) } =\frac { 1 }{ 2 } } \\ $$

- 21,578
The first couple of terms of the Euler-Maclaurin Sum Formula can be derived as
$$
\begin{align}
\sum_{k=1}^nk^p
&=\int_{0^+}^{n^+}x^p\,\mathrm{d}\!\left\lfloor x\right\rfloor\tag1\\
&=\int_{0^+}^{n^+}x^p\,\mathrm{d}(x-\{x\})\tag2\\
&=\tfrac1{p+1}n^{p+1}-\int_{0^+}^{n^+}x^p\,\mathrm{d}\!\left(\{x\}-\tfrac12\right)\tag3\\
&=\tfrac1{p+1}n^{p+1}-\left[x^p\left(\{x\}-\tfrac12\right)\right]_{0^+}^{n^+}+p\int_0^n\left(\{x\}-\tfrac12\right)x^{p-1}\,\mathrm{d}x\tag4\\[6pt]
&=\tfrac1{p+1}n^{p+1}+\tfrac12n^p+O\!\left(n^{p-1}\right)\tag5
\end{align}
$$
Explanation:
$(1)$: write sum as a Riemann-Stieltjes Integral
$(2)$: $\left\lfloor x\right\rfloor=x-\{x\}$
$(3)$: use $\{x\}-\frac12$ because its integral is periodic
$(4)$: Integrate by Parts
$(5)$: see below
Proof of $(5)$:
$$
\begin{align}
\left|\,p\int_0^n\left(\{x\}-\tfrac12\right)x^{p-1}\,\mathrm{d}x\,\right|
&=\left|\,p\int_0^n\left(\{x\}-\tfrac12\right)\left(x^{p-1}-\left\lfloor x\right\rfloor^{p-1}\right)\,\mathrm{d}x\,\right|\tag6\\
&=\left|\,p\sum_{k=0}^{n-1}\int_0^1\left(x-\tfrac12\right)\left((k+x)^{p-1}-k^{p-1}\right)\,\mathrm{d}x\,\right|\tag7\\
&\le p\sum_{k=0}^{n-1}\tfrac12\left((k+1)^{p-1}-k^{p-1}\right)\tag8\\[6pt]
&=\tfrac{p}2n^{p-1}\tag9
\end{align}
$$
Explanation:
$(6)$: the integral of $\{x\}-\tfrac12$ over a unit interval is $0$
$(7)$: for $(k,x)\in\mathbb{Z}\times[0,1)$, $\{k+x\}=x$ and $\lfloor k+x\rfloor=x$
$(8)$: for $(k,x)\in\mathbb{Z}\times[0,1)$, $\left|x-\frac12\right|\le\frac12$ and $\left|(k+x)^{p-1}-k^{p-1}\right|\le(k+1)^{p-1}-k^{p-1}$
$(9)$: Telescoping Sum
Divide $(5)$ by $n^p$, subtract $\frac{n}{p+1}$, and take the limit.

- 345,667
-
-
@user1357113: getting my uncle's estate taken care of. Trying to keep my cold from extending into Thanksgiving Day (this Thursday). I hope things are well with you. – robjohn Nov 25 '19 at 14:28
-
-
-
@user1357113: haven't had a chance to look at it, but I know the UCLA Library has access to the electronic edition. – robjohn Nov 25 '19 at 20:09
This is just to show an alternative proof of the fact stated by @Did as I was not able to follow his argument completely (not for lack of clarity on Did's part but of my lack of mental capacity I suppose).
Lebesgue integration by parts yields $$ \sum_{a<k\leq b}g(k)=\int^b_a g(x)\,dx +\int^b_a g'(x)\{x\}\,dx+g(a)\{a\}-g(b)\{b\} $$ for any absolutely continuous function $g$ on $[a,b]$, where $\{x\}=x-\lfloor x\rfloor$ is the fractional part function.
Letting $g(x)=f(x/n)$ for $0\leq x\leq n$ we obtain \begin{align} \sum_{0<k\leq n}f(k/n)&=\int^n_0 f(x/n)\,dx + \frac1n\int^n_0f'(x/n)\{x\}\,dx\\ &=n\int^1_0f(x)\,dx +\int^1_0 f'(x)\{nx\}\,dx \end{align}
Now, the function $x\mapsto\{x\}$ is 1-periodic. An application of Fejer's lemma yields $$\lim_{n\rightarrow\infty}\int^1_0 f'(x)\{nx\}\,dx=\Big(\int^1_0\{x\}\,dx\Big)\int^1_0f'(x)\,dx=\frac{f(1)-f(0)}{2}$$
Hence $$\sum^n_{k=1}f(k/n) - n\int^1_0f(x)\,dx\xrightarrow{n\rightarrow\infty}\frac{f(1)-f(0)}{2}$$

- 39,145
-
+1 nice proof. I knew the result which you call Fejer lemma but didn't know it was famous by that name. – Paramanand Singh Oct 31 '22 at 17:38
-
@ParamanandSingh: This is just a minor result of his on his vast work on Fourier analysis. – Mittens Oct 31 '22 at 17:40
Hint for another method:
You can use Faulhaber's formula: $$1^p+2^p+\dots+n^p=\frac1{p+1}\sum_{k=0}^p\binom{p+1}kB_k\, n^{p+1-k},$$ where $B_k$ denotes the $k$-th Bernoulli number (with the convention that $B_1=-\frac12$).

- 175,478