For the random variables X, Y with joint density function

$f(x,y)=cx^2y(1-x)(y-1) \space, \space \space \space 0 \leq x \leq 1,\space \space0 \leq y \leq 1$$

$f(x,y)=0 ,\space \space \space \space otherwise$

(a) For what value of c is this a joint density function?

(b) Using this value of c, compute the density functions of X and Y .

I'm not sure of my answer in part a), I took the double integral as such: $$\int_{0}^{1} \int_{0}^{1} cx^2y(1-x)(y-1)dxdy = 1$$

Am I right by doing it this way?

Also if I understand part b) correctly, am I right to say that I would need to compute the following: $$\int_{0}^{1} \int_{0}^{1} -72x^2y(1-x)(y-1)dxdy $$

  • 1
    $\begingroup$ For part $a)$ what you are doing sounds correct, you want to integrate $f(x,y)$ over the domain and you will get some value $Z$ say, now if $Z = 1$ you are done, else take $c = 1/Z$. For part $b)$ I understand that to mean calculate the marginal densities of $X$ and $Y$, that is to say from a new function $f(x) = \int c f(x,y) dy$ and if necessary rescale so that $f(x)$ is a valid pdf, and do the same for $f(y)$ $\endgroup$ – Nadiels Feb 19 '17 at 22:39
  • $\begingroup$ I don't quite get it for part b)... is my reasoning ok above? $\endgroup$ – JKawa Feb 20 '17 at 1:22
  • 1
    $\begingroup$ No. You just want: $f_X(x) = \int_0^1 -72x^2 y(1-x)(y-1)\operatorname d y = 12x^2(1-x)$ (when $x\in(0;1)$) and similarly for $f_Y(y)$,,, $\endgroup$ – Graham Kemp Feb 20 '17 at 2:28

Part a is correct.

Part b:

$ f_X(x) = \int f(x,y)\operatorname d y = \int_0^1 -72x^2 y(1-x)(y-1)\operatorname d y = 12x^2(1-x) $ and $ f_Y(y) = \int f(x,y)\operatorname d x = \int_0^1 -72x^2 y(1-x)(y-1)\operatorname d x = -6y(y-1). $

Moreover $X$ and $Y$ are independent since $f(x,y) = f_X(x)f_Y(y).$

| 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.