Let $(X_i)_{i\in\mathbb{N}}$ be a strictly stationary sequences of real valued random variables with finite variance. We have the empirical distribution functions $F_{n}(u):=\frac{1}{n} \sum_{i=1}^n 1\{X_i\leq u\}$ and a continuous distribution function $F$. Assume that for all $u\in\mathbb{R}$ \begin{align*} |F_{n}(u)-F(u)|\xrightarrow{p} 0 \end{align*}in probability for $n\rightarrow\infty$.
Holds now that \begin{align*} \sup_{u\in\mathbb{R}} |F_{n}(u)-F(u)|\xrightarrow{p} 0? \end{align*} Thanks!

  • 1
    $\begingroup$ The answer is yes if $(X_i,i\in\Bbb N)$ is ergodic. It's a result from Dehling and Philipp, 2002. $\endgroup$ – Davide Giraudo Jul 11 '13 at 13:46

We can use Corollary 1.4 of paper

Ramon van Handel, The universal Glivenko-Cantelli property, Probab. Th. Rel. Fields 155, 911-934 (2013)

which is available here (especially point 8). The assumption about pointwise convergence allows us to see that for all $u$,
$$F(u)=\mathbb P(\chi_{\{Z_0\leqslant u\}}\mid\mathcal I)\quad\mbox{p.s}.$$


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.