Here's an interesting question:

The area of a circle is exponentially distributed with parameter $\lambda$. Find the distribution of PDF of the radius of the circle. Then assume that the radius is exponentially distributed (with the same rate $\lambda$) and find the PDF of the circle’s area.

So an exponential distribution means that the density function would be along the lines of:

PDF(x) = f(x) = $\lambda e^{-\lambda x}dx$

then I think one would do something along these lines:

$P${$R<=r$} = $ \frac{1}{2\pi} \int_{B(0)}\int r^\frac{r^2}{2}dr$ = $\int^0_r r^{\frac{r^2}{2}} dr$

to find the PDF of the radius, right?

How would I apply this to find the PDF of the circle's area?

  • 1
    $\begingroup$ You would not apply this, they are two separate problems. The integral is not right. For one thing, where is the exponential distribution? $\endgroup$ – André Nicolas May 1 '12 at 2:56
  • $\begingroup$ I think I see what you were saying, and I think I fixed my integral... What else am I doing wrong here? I've been breaking my head open with this one for a while. $\endgroup$ – gfppaste May 1 '12 at 3:00
  • $\begingroup$ I posted a solution for the first problem, and a start on the second, since you seem to have real difficulties. There may be places in the first problem where you need extra help. I will not give extra help on the second, except to confirm whether your answer is right. $\endgroup$ – André Nicolas May 1 '12 at 3:15
  • $\begingroup$ Thank you very much for your help on this one... you've made the problem make a lot more sense. $\endgroup$ – gfppaste May 1 '12 at 3:20

Let the random variable $X$ denote the area, and let the random variable $R$ denote the radius. We have been told that the density function of the random variable $X$ is $\lambda e^{-\lambda x}$ for $x\ge 0$, and $0$ elsewhere. That is what is meant by saying that area has exponential distribution with parameter $\lambda$.

We want the pdf of $R$. To get it, we first find an expression for the cumulative distribution function of $R$, that is, the probability that $R\le r$. This is $0$ for $r\lt 0$, so assume $r\ge 0$. We have $X=\pi R^2$. So $R \le r$ iff $X \le \pi r^2$, and therefore $$P(R\le r)=P(X\le \pi r^2)=\int_0^{\pi r^2} \lambda e^{-\lambda x}\,dx.$$ We could integrate, it is not hard, and then differentiate to get the density function of $R$. But it is easier (at least for me) to differentiate under the integral sign, remembering to use the Chain Rule. We get that (for $r\ge 0$) $$f_R(r)= (2\pi r)\lambda e^{-\lambda \pi r^2}.$$

Now for the other problem. Here it is $R$ that has the exponential distribution. We want to find $P(X\le x)$. Since $X=\pi R^2$, this is true iff $R\le \sqrt{x/\pi}$. Express this as an integral, in the style of the first part. Then evaluate the integral and differentiate, or differentiate without evaluating.

| cite | improve this answer | |

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.