3
$\begingroup$

Suppose $x_1, x_2$ are IDD random variables uniformly distributed on the interval $(0,1)$. What is the pdf of the quotient $x_2 / x_1$?

$\endgroup$
  • $\begingroup$ Duplicate of this ? $\endgroup$ – Sasha Feb 25 '12 at 18:37
7
$\begingroup$

Because $X_2$ is positive with almost surely, cumulative distribution function for $Z=X_1/X_2$ is $$ F_Z(z) = \mathbb{P}(Z \leqslant z) = \mathbb{P}(X_1 \leqslant z X_2) = \mathbb{E}_{X_2}\left( \mathbb{P}(X_1 \leqslant z X_2 | X_2)\right) = \mathbb{E}_{X_2}\left( F_{X_1}(z X_2)\right) $$ Clearly $F_Z(z)=0$ for $z\leqslant 0$, so assume $z > 0$ $$ F_Z(z) = \int_0^1 \left\{ \begin{array}{cl} z x_2 & 0 < x_2 <1/z \\ 1 & x_2 > 1/z \end{array} \right. \mathrm{d} x_2 = \frac{z}{2} \left(\frac{1}{\max(z,1)}\right)^2 + \left(1 - \frac{1}{\max(z,1)}\right) = \left\{\begin{array}{cl} \frac{z}{2} & 0< z\leqslant 1 \\ 1-\frac{1}{2z} & z > 1 \end{array} \right. $$ The probability density is obtained by differentiation.

$\endgroup$
  • 5
    $\begingroup$ You could also draw what is happening in the [0,1]x[0,1] square and identify z/2 and 1/(2z) as the areas of some simple triangles. $\endgroup$ – Did Feb 25 '12 at 19:01

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.