How to find bounds of integration for this question?

49 Views Asked by At

I am currently studying joint pdfs and am struggling with finding the bounds of integrations for this problem. Consider a pair of random variables X, Y with constant joint density on the triangle with vertices at $(0, 0)$, $(3, 0)$, and $(0, 3)$. Find $P(X +Y > 2)$.

The fact that it says constant density tells me that this is continuous uniform with area $0.5 \cdot 3 \cdot 3 = 4.5$ so $f(x) = 2/9$

Since $P(X +Y > 2)$ I thought that this would mean that the bounds of integration would be $\displaystyle\int_2^3\int_{2-x}^3\cfrac{2}{9}dydx = \cfrac{7}{9}$

But my book says the answer is $\cfrac{5}{9}$

What am I misunderstanding here? What are the proper bounds of integration and why?

3

There are 3 best solutions below

0
On BEST ANSWER

If you are using integral, then notice it is the shaded area you need to integrate over. So easier is to integrate over the unshaded part and subtract from $1$.

enter image description here

So the integral should be

$P(X+Y \lt 2) = \displaystyle\int_0^2\int_{0}^{2-y} \cfrac{2}{9} \ dx \ dy = \cfrac{4}{9}$

So $P(X+Y \gt 2) = \displaystyle 1 - \frac{4}{9} = \frac{5}{9}$.

0
On

The endpoints of the integral are wrong in a few places: it should be $$ \int_0^3\int_{\max\{2-x,0\}}^{3-x}\cfrac{2}{9} \, dy \, dx = \frac59. $$

  • When you put the lower limit of the first integral equal to $2$, that prohibited $X$ from being less than $2$;
  • Putting $2-x$ as the lower limit of the second integral is an error because $2-x$ can be negative, and $f(x,y)=0$ for $y<0$;
  • Having $3$ as the upper limit of the second integral allowed pairs $(x,y)$ with $x+y>3$, contrary to the initial domain of the random variables.
0
On

A direct calculation can be made by observing that the entire domain is an isosceles right triangle with side length $3$ so the area is $\frac{9}{2}$. The condition is exclusion of a similar right triangle with side length $2$ and area$\frac{4}{2}$ leaving the area for $P(X+Y\gt 2)$ as $\frac{5}{2}$ and a probability of $\frac{5}{9}$.