# sum of two dependent random variables

Let $X$ be a cotinuous random variable uniformly distributed over $[-10,10]$. Let $Y$ be a random variable with pdf $f_Y(y) = \frac{1}{40}\ln \frac{20}{|y|}, -20 \leq y \leq 20$.

$X$ and $Y$ ARE NOT independent. How do find the pdf of the sum $X+Y$?

I can't use convolution because of their non-independence.

How do I begin?

• How do you know that X and Y are not independent ? What is the link between X and Y ? – Dark Jan 27 '15 at 19:15
• before getting the pdf of $Y$, $Y$ is actually a function of $X$. Is there a way to check in/dependence? – cgo Jan 27 '15 at 19:18
• What function then? Give all information you have. – drhab Jan 27 '15 at 19:22
• So we want to find the sum $X+Y$, where $Y=2Z(c-X)$. $2Z \sim U(0,2)$ and $X\sim U(-10,10)$ as mentioned above. – cgo Jan 27 '15 at 19:27
• You need to know what the dependence is exactly. That is, you need to know the joint distribution – Luis Mendo Jan 28 '15 at 16:35

You have that : $Y = f(X)$ and you want the pdf of $X+Y$.
Let $\phi : \mathbb{R}^2 \to \mathbb{R}$ be a smooth bounded function.
Compute $\mathbb{E}[\phi(X+Y,X)]$. That should give you the joint pdf of $(X+Y,X)$. Then you should know how to get the pdf of $X+Y$.