Here is the problem I meet in analytical number theory($p$ is prime number here):
Prove: $$ \sum_{p \leq N} \ln^2 p= N\ln N + o(N\ln N) $$
I really don't know where to start or what is the right theorem to use. Any help is appreciated!
Here is the problem I meet in analytical number theory($p$ is prime number here):
Prove: $$ \sum_{p \leq N} \ln^2 p= N\ln N + o(N\ln N) $$
I really don't know where to start or what is the right theorem to use. Any help is appreciated!
On
Note that we can find more terms of this asymptotic quite easily. From the PNT in the form $$\pi\left(x\right)=\frac{x}{\log\left(x\right)}\sum_{k=0}^{n}\frac{k!}{\log^{k}\left(x\right)}+O_{n}\left(\frac{x}{\log^{n+2}\left(x\right)}\right)$$ we have, taking for example $n=1$ and using the Abel summation formula, that $$\sum_{p\leq N}\log^{2}\left(p\right)=\pi\left(N\right)\log^{2}\left(N\right)-2\int_{2}^{N}\frac{\pi\left(t\right)\log\left(t\right)}{t}dt$$ $$=N\log\left(N\right)+N+O\left(\frac{N}{\log\left(N\right)}\right)-2\int_{2}^{N}\left(1+\frac{1}{\log\left(t\right)}\right)dt+O\left(\int_{2}^{N}\frac{1}{\log^{2}\left(t\right)}dt\right).$$ Now, integrating by parts, it is not difficult to see that $$\int_{2}^{N}\frac{1}{\log\left(t\right)}dt\sim\frac{N}{\log\left(N\right)}$$ and $$\int_{2}^{N}\frac{1}{\log^{2}\left(t\right)}dt\ll\int_{2}^{N}\frac{1}{\log\left(t\right)}dt\sim\frac{N}{\log\left(N\right)}$$ so $$\sum_{p\leq N}\log^{2}\left(p\right)=\color{red}{N\log\left(N\right)-N+O\left(\frac{N}{\log\left(N\right)}\right)}.$$
Let $\mathbf{1}_p(n)$ the function which equals $1$ if $n$ is a prime and $0$ otherwise. By summation by parts
$$ \sum_{p\leq N}\log^2 p = \sum_{n\leq N}\mathbf{1}_p(n) \log^2 n = \pi(N)\log^2 N-\sum_{n\leq N-1} \pi(n)\left[\log^2(n+1)-\log^2(n)\right] $$ On the other hand $\pi(N)\sim\frac{N}{\log N}$ by the PNT and $$ \log^2(n+1)-\log^2(n)=(\log n+\log(n+1))\log\left(1+\frac{1}{n}\right) = O\left(\frac{\log n}{n}\right) $$ such that $$\sum_{n\leq N}\pi(n)\left[\log^2(n+1)+\log^2(n)\right]\ll\sum_{n\leq N}1 = O(N)$$ and the claim is proved. Actually a stronger version of it, where $o(N\log N)$ has been replaced by $O(N)$.