I tried to work on the problem

Let $(X_n)$ be a sequence of random variables follows the discrete uniform distribution on $\{0,\frac{1}{n},\frac{2}{n},\cdots,\frac{n-1}{n}\}$. Let $X$ be a continuous uniform random variable on $(0,1)$.

I want to show that $X_n$ converges weakly to $X$. That is

$E[f(X_n)] \rightarrow E[f(X)]$ for all bounded and continuous function $f$.

I know that $X_n$ converges to $X$ in distribution, $X_n \overset{D}{\rightarrow} X$ and this implies the weak convergence. But how can I just use the definition from above to show this type of convergence?


1 Answer 1


By the definition of expectation:
\begin{align*} & E[f(X_n)] = \sum_{i = 0}^{n - 1}f(i/n)n^{-1} = \sum_{i = 0}^{n - 1}\int_{i/n}^{(i + 1)/n}f(i/n)dx, \\ & E[f(X)] = \int_0^1 f(x)dx = \sum_{i = 0}^{n - 1}\int_{i/n}^{(i + 1)/n}f(x)dx. \end{align*}

Since $f$ is continuous on $\mathbb{R}$, it is uniformly continuous on $[0, 1]$. That means, given $\varepsilon > 0$, there exists $\delta > 0$, such that for all $x, y \in [0, 1]$ and $|x - y| < \delta$, we have $|f(x) - f(y)| < \varepsilon$. Let $N \in \mathbb{N}$ be sufficiently large so that $\frac{1}{N} < \delta$, then for all $n > N$, $x \in [i/n, (i + 1)/n)$ implies that $|x - i/n| < \frac{1}{n} < \frac{1}{N} < \delta$, whence \begin{align} |E[f(X_n)] - E[f(X)]| \leq \sum_{i = 0}^{n - 1}\int_{i/n}^{(i + 1)/n}|f(i/n) - f(x)|dx < \sum_{i = 0}^{n - 1}\varepsilon \times \frac{1}{n} = \varepsilon. \end{align} This shows $E[f(X_n)] \to E[f(X)]$ as $n \to \infty$.

Note that the boundedness condition is actually not needed for this particular case because the support of the uniform distribution is a compact set (it is, of course, still essential in proving the equivalence between weak convergence and convergence in distribution for general random variables).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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