In this vignette, we’ll present a real-life use case, which shows how the matricks package makes the work with matrices easier.

Let’s try to implement an algorithm from the field of Reinforcement Learning called iterative policy evaluation. The environment we will work on is a simple Grid World game. This vignette was inspired by a Reinforcement Learning course by Lazy Programmer, originally written in Python.


Initializing environment

We initialize environment, using following functions from matrix package: * m() * matrix_idx() * plot.matrix() * seq_matrix() * sv() * with_same_dims()

We construct matrices, describing game environment: * actions - all the possible moves * rewards