1
$\begingroup$

The Weibull pdf is given as the following.

$$f(x) = \begin{cases} \frac{1}{\alpha}mx^{m-1}e^{-x^m/\alpha} \quad \text{if } x > 0 \\ 0 \quad \text{else} \end{cases}$$

If a random sample of size $n$ is taken from a Weibull distribution, what is the pdf and cdf of the minimum of the sample, $Y_i$?

I realize $g_1(y_1) = n[1-F(y_1)]^{n-1}f(y_1)$; however how do I find the cdf $F(y_1)$? I tried integrating the pdf from 0 to $y$ but am having difficulty.

$\endgroup$
2
$\begingroup$

Let $X_1,X_2,\dots,X_n$ be iid Weibull, and let $W$ be the minimum of the $X_i$.

Then $W\gt w$ if and only if all the $X_i$ are $\gt w$. This has probability $(1-F_X(w))^n$, where $F_X$ is the cdf of one of the Weibulls. Thus the cdf of $W$ is given by $$F_W(w)=1-\Pr(W\gt w)=1-(1-F_X(w))^n$$ (for positive $w$).

To find $1-F_X(w)$, we need to calculate $$\int_w^\infty \frac{1}{\alpha}mx^{m-1}\exp(-x^m/\alpha)\,dx.$$ Make the substitution $\frac{x^m}{\alpha}=t$. Then $dt=\frac{1}{\alpha}mx^{m-1}\,dx$, and the integral becomes the easy $$\int_{t=w^m/\alpha}^\infty e^{-t}\,dt.$$

$\endgroup$
  • $\begingroup$ How would I find the pdf of the minimum of the sample? $\endgroup$ – idealistikz Jun 13 '13 at 6:29
  • $\begingroup$ I have shown you how to find the cdf (cumulative distribution function). To find the probability density function, differentiate. We get $nf_X(w)(1-F_X(w))^{n-1}$. But your post seemed to indicate you knew that. $\endgroup$ – André Nicolas Jun 13 '13 at 6:42

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.