Conditional Distributions: Discrete Case
What you'll learn: How to compute the probability distribution of one discrete random variable when another discrete random variable takes a specific value.
The Core Idea
Imagine you and a friend are rolling dice together. Knowing your friend's roll changes what you'd expect about the total sum, right? That's conditioning in action.
When we have two discrete random variables and with a joint PMF, we often want to know: "What's the distribution of given that equals some specific value ?" This gives us the conditional probability mass function (conditional PMF).
The Conditional PMF Formula
The conditional PMF of given is:
provided that .
In plain English: Take the joint probability that both events occur, and divide by the marginal probability of the conditioning event. This normalizes the probabilities so they sum to 1 over all possible values.
How It Works
- Start with the joint PMF for all pairs
- Find the marginal PMF by summing the joint PMF over all values
- Divide the joint by the marginal for your specific value
- Verify that (it's a valid PMF)
This creates a "slice" of the joint distribution, rescaled to be a proper probability distribution on its own.
Key Takeaway: The conditional PMF describes the probability distribution of when is held fixed at a specific value, found by normalizing the joint probabilities.