The Python Interface

The Python environment that will help facilitate our discussions of MAB algorithms is included in the utils.py file of this chapter's code repository on GitHub: https://packt.live/3cWiZ8j.

From this file, we can import the Bandit class into a separate script or a Jupyter script. This class is the interface we will use to create, interact, and solve various MAB problems. If the code we are working with is in the same directory as this file, we can import the Bandit class by simply using the following code:

from utils import Bandit

Then, we can declare an MAB problem as an instance of a Bandit object:

my_bandit = Bandit()

Since we are not passing any arguments to this declaration, this Bandit instance takes on its default value: an MAB problem with two Bernoulli arms with probabilities of 0.7 and 0.3 (although our algorithms technically cannot know this).

The most integral method of the Bandit class that we need to be aware of is pull(). This...