I know that for marginal distributions $f(x)$ with cumulative distribution $F(x)$, we derive $f(x) = d F(x)/dx$

My question: If given a simultaneous cumulative probability density $F(x_1,x_2,\cdots,x_n)$ how can one derive $f(x_1,x_2,\cdots,x_n)$?

  • 1
    $\begingroup$ As written in Wikipedia, the pdf is the $n$-th derivative of the cdf against all its components. $\endgroup$ – Xi'an Apr 20 '17 at 19:55

It's just as you'd expect: the $cdf$ is differentiated with repect to each of the random variables. That is for a $cdf$ of n random variables $x_1,...,x_n$: $$f(x_1,x_2,...,x_n) = \frac{∂^n}{∂x_1 ...∂x_n} F(x_1,x_2,...,x_n)$$

However this must assume that the $cdf$ is continuous and the partial derivatives exist.

Some nice reference are: here and here

| 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.