If $X$ has a $\Gamma(\alpha, 1)$ distribution would $\beta X$ not have a $\Gamma(\alpha \beta, 1)$ distribution by the summation property of Gamma distributions? I must be missing something.

  • 6
    $\begingroup$ you're not summing. You're scaling. You can prove both with moment generating functions, though. $\endgroup$
    – Taylor
    Nov 30 '17 at 3:14

Taylor's suggestion to use mgf is good. I show you how to find the distribution of $\beta X$ by transformation.

$X\sim \Gamma(\alpha,1) \Rightarrow f_{X}(x)=\frac{1}{\Gamma(\alpha)}x^{\alpha-1}e^{-x}$

Let $Y=\beta X\Rightarrow x=\frac{1}{\beta}y$

$\therefore |J|=\frac{\partial x}{\partial y}=\frac{1}{\beta}$

$ f_{Y}(y)=f_{X}(x)|J|=\frac{1}{\Gamma(\alpha)}(\frac{1}{\beta}y)^{\alpha-1}e^{-\frac{1}{\beta}y}\frac{1}{\beta}=\frac{1}{\Gamma(\alpha)\beta^{\alpha}}y^{\alpha-1}e^{-\frac{y}{\beta}}$

This is the pdf of $\Gamma(\alpha,\beta)$

$\therefore Y\sim \Gamma(\alpha, \beta)$

It is not $\Gamma(\alpha\beta,1)$ as you have thought


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.