- joint CDF: probability both variables are at or below a point, the two variable version of the CDF
FX,Y(x,y)=P(X≤x∩Y≤y)
- from a joint mass, add up every cell at or below (x,y)
| y=0 | y=1 |
|---|
| x=0 | 103 | 103 |
| x=1 | 102 | 102 |
- $F(0,0) = P(X=0 \cap Y=0) = \frac{3}{10}$
- $F(0,1) = P(X=0 \cap Y=0) + P(X=0 \cap Y=1) = \frac{6}{10}$
- same for the rest, giving the joint CDF
| y=0 | y=1 |
|---|
| x=0 | 103 | 106 |
| x=1 | 105 | 1010 |
- marginal CDF: send the other variable to infinity, i.e. stop constraining it
FX(x)=y→∞limFX,Y(x,y),FY(y)=x→∞limFX,Y(x,y)
- here $y=1$ is already the top, so $F_X(0) = F(0,1) = \frac{6}{10}$ and $F_X(1) = F(1,1) = 1$
- and $F_Y(0) = F(1,0) = \frac{5}{10}$, $F_Y(1) = 1$
- independent variables factor in the CDF just like in the mass
- pX,Y(x,y)=pX(x)pY(y) and FX,Y(x,y)=FX(x)FY(y)
- check: F(0,0)=103=106⋅105
- equivalently conditioning changes nothing: pX∣Y(x∣y)=pX(x), pY∣X(y∣x)=pY(y)