Can I say that the random variables are independent by looking at the joint pdf and iff I can factor the two random variables from each other, say (X+1)(Y+2), they are independent?

If not, is there another way one can use to quickly determine based on the joint pdf that two RV's are independent?

Of do I always have to integrate out the other var to find the other marginal pdfs, multiply them together and see if they equal the joint?

  • 3
    $\begingroup$ If you have a joint density that factors, you have independence. But one has to be careful, it has to be the whole density function, including where it is $0$. or if you stick to non-zero part, want it over a "rectangle" (possibly infinite) with sides parallel to the axes. $\endgroup$ – André Nicolas Oct 4 '12 at 5:09

If you can factor the joint pdf $f(x,y)$ into $g(x)h(y)$ then $X$ and $Y$ are indeed independent.

The marginal pdfs are $\dfrac{g(x)}{\int_x g(x) \,dx}$ and $\dfrac{h(y)}{\int_y h(y) \,dy}$ while $\left(\int_x g(x) \,dx\right )\left(\int_y h(y) \,dy\right)=\int_x \int_y g(x)\,h(y)\, dy \,dx=\int_x \int_y f(x,y) \,dy\, dx=1$.


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.