Let $X$ be a random variable from a chi-square distribution with 1 degree of freedom. The probability density function (pdf) of $X$ is $f(x) = \frac{\exp{(-x/2)}}{\sqrt{2\pi x}}$, $x>0$.

In the case of half-chi-square distribution with 1 degree of freedom ($\frac{1}{2} \chi^{2}_{1}$), is the pdf provided by $\frac{1}{2}\frac{\exp{(-x/2)}}{\sqrt{2\pi x}}$?


Here's a useful general result.

Theorem. Suppose $X$ is a random variable with cumulative distribution function $F_X$, let $a, b \in \mathbb{R}$ with $b > 0$, and let $Y = a + b X$.

  1. The cumulative distribution function of $Y$ (call it $F_Y$) is given by $$ F_Y(y) = F_X\left(\frac{y - a}{b}\right) $$ for all $y \in \mathbb{R}$.
  2. If $X$ is also continuous with density $f_X$, then the density of $Y$ (call it $f_Y$) is given by $$ f_Y(y) = \frac{1}{b} f_X\left(\frac{y - a}{b}\right) $$ for all $y \in \mathbb{R}$.

Your question is the special case where $X \sim \chi^2_1$ (whose density you know), and you're interested in the density of $Y = \frac{1}{2} X$ (here $a = 0$ and $b = 1/2$). The theorem above will allow you to compute the density of $Y$.

Proof of 1. Just compute: $$ \begin{aligned} F_Y(y) &= P(Y \leq y) \\ &= P(a + b X \leq y) \\ &= P\left(X \leq \frac{y - a}{b}\right) && \text{(*)}\\ &= F_X\left(\frac{y - a}{b}\right). \end{aligned} $$ (*) Here we used the assumption that $b > 0$. If $b < 0$, the inequality would change direction, and if $b = 0$, we wouldn't be able to divide by $b$ at all.

Proof of 2. Differentiating the result of the first part, we get $$ \begin{aligned} f_Y(y) &= \frac{d}{dy} F_Y(y) &&\text{(**)}\\ &= \frac{d}{dy} F_X\left(\frac{y - a}{b}\right) \\ &= \frac{1}{b} F^\prime_X\left(\frac{y - a}{b}\right) &&\text{(chain rule)}\\ &= \frac{1}{b} f_X\left(\frac{y - a}{b}\right). &&\text{(**)} \end{aligned} $$ (**) Here we used the fact that a density of a continuous random variable is the derivative of that random variable's cumulative distribution function, which is a consequence of the definition of the density and the fundamental theorem of calculus.

  • 1
    $\begingroup$ You might want to re-emphasize the $b>0$ assumption since it is critical to both parts 1. and 2. $\endgroup$ Feb 8 '19 at 11:29
  • $\begingroup$ @DilipSarwate good suggestion; I've updated my answer $\endgroup$ Feb 8 '19 at 17:40

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.