I'm a bit of a novice at maths and want to learn more about algebraically manipulating likelihoods in statistics. There are a lot of equations that involve taking the product of a set of values given a model.
I know a few rules for manipulating sigma-notation (e.g., here and here).
- What are the basic rules for manipulating sequences of products (i.e, $\prod_{i=1}^{I} ... $)? Is there a web page that you could direct me to?
e.g.,
- $\prod_{i=1}^{I} x_i$
- $\prod_{i=1}^{I} x_i y_i$
- $\prod_{i=1}^{I} a + b x_i$
- $\prod_{i=1}^{I} \exp x_i$
You just need to think about what the product notation means, and you can work out the rules yourself.
Take one of your examples:
$$\prod_{i=1}^I x_i y_i = \left(\prod_{i=1}^I x_i\right)\left(\prod_{i=1}^I y_i\right)$$
This rule works because the left side multiplies $x_1$ times $y_1$ times $x_2$ times $y_2$ etc., and the right side multiplies the x's first, then multiplies the result by the y's. Clearly those are equal.
Another one of your examples:
$$\prod_{i=1}^I \exp x_i = \exp \left( \sum_{i=1}^I x_i \right)$$
Can you figure out why this rule works?