0
$\begingroup$

We are given a distribution, $\mathbb{P}(Y_n=\frac k n)=2^{-k}$ for $k= 1, 2, 3,\dots$. Check if this converges in distribution and if it does find the limit distribution.

I think it does, the CDF function for $Y_n$ looks as follows (?):

$$F_n=\begin{cases} 0 &nt<1 \\ \sum_{k=1}^{\lfloor nt \rfloor} 2^{-k} &nt\geq1 \end{cases}$$

in the limit we get $$F_n\to\begin{cases} 0 & t\leq0 \\ 1 &t>0 \end{cases}$$

This is not a distribution because it is not right continous however $$F(t)=\begin{cases} 0 & t<0 \\ 1 &t\geq0 \end{cases}$$

is a distribution and $F_n$ converges to $F$ in points of continuity of $F$ right?

So indeed $F_n$ converges.

$\endgroup$
  • 1
    $\begingroup$ in the first $\sum_{k=1}^{\lfloor t \rfloor} 2^{-k}$ you should replace $t$ by $nt$. The rest is wrong because of this mistake. $\endgroup$ – justt Dec 5 '19 at 18:05
  • $\begingroup$ @justt what about now? $\endgroup$ – Kran Dec 5 '19 at 18:51
0
$\begingroup$

It looks like you took a limit improperly. Try this hint instead:

Hint: Fix $\epsilon > 0$ and calculate $\mathbb{P}(Y_n > \epsilon)$. What happens as $n \to \infty$?

$\endgroup$
  • $\begingroup$ I typed it wrong! I did not even considered your hint, but I think it is alright now. $\endgroup$ – Kran Dec 5 '19 at 18:13
  • $\begingroup$ No in fact it was really wrong, however now after another editing it is right i think. $\endgroup$ – Kran Dec 5 '19 at 18:16

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.