Chapter 27: Conditional Distributions

Meike Niederhausen and Nicky Wakim

2023-11-15

Learning Objectives

  1. Calculate the conditional probability density from a joint pdf

Conditional probabilities we’ve seen before

What do we know about conditional probabilities for events and discrete RVs?    

For events:

\[P(A | B) = \dfrac{P(A \cap B)}{P(B)}\]    

For discrete RVs: \[p_{X|Y}(x|y) = P(X=x|Y=y) = \dfrac{p_{X,Y}(x,y)}{p_Y(y)}\]

What does it mean for conditional densities of continuous RVs?

For continuous RVs:

Example starting from a joint pdf: first try!

Example 1.1

Let \(f_{X,Y}(x,y)= 5 e^{-x-3y}\), for \(0 < y < \frac{x}{2}\).

  1. Find \(\mathbb{P}(2<X<10|Y=4)\)

What is a conditional density?

Definition: Conditional density

The conditional density of a r.v. \(X\) given \(Y=y\), is \[f_{X|Y}(x|y)= \frac{f_{X,Y}(x,y)}{f_Y(y)},\] for \(f_Y(y)> 0\)

Remarks

  1. It follows from the definition for the conditional density \(f_{X|Y}(x|y)\), that \[f_{X,Y}(x,y)= f_{X|Y}(x|y)f_Y(y).\]

     

  1. For a fixed value of \(Y=y\), the conditional density \(f_{X|Y}(x|y)\) is an actual pdf, meaning

    • \(f_{X|Y}(x|y)\geq 0\) for all \(x\) and \(y\), and

    • \(\displaystyle\int_{-\infty}^{\infty} f_{X|Y}(x|y)dx =1\).

Example starting from a joint pdf: second try!

Example 1.1

Let \(f_{X,Y}(x,y)= 5 e^{-x-3y}\), for \(0 < y < \frac{x}{2}\).

  1. Find \(\mathbb{P}(2<X<10|Y=4)\)

Example starting from a joint pdf

Example 1.2

Let \(f_{X,Y}(x,y)= 5 e^{-x-3y}\), for \(0 < y < \frac{x}{2}\).

  1. Find \(\mathbb{P}(X>20 |Y=5)\)

Finding probability with conditional domain and pdf

Example 2

Randomly choose a point \(X\) from the interval \([0,1]\), and given \(X=x\), randomly choose a point \(Y\) from \([0,x]\). Find \(\mathbb{P}(0 < Y < \frac14)\).

Independence and conditional distributions

    Question What is \(f_{X|Y}(x|y)\) if \(X\) and \(Y\) are independent?

\[f_{X|Y}(x|y) = \dfrac{f_{X,Y}(x,y)}{f_y(y)} = \dfrac{f_{X}(x)f_y(y)}{f_y(y)} = f_{X}(x)\]

       

  • If \(f_{X|Y}(x|y)\) does not depend on \(y\) (including the bounds/domain), then \(X\) and \(Y\) are independent.