In this one, I try to make custom environments for a rl algo to learn in.
1 . PDE (pickup-delivey-environment) code
PDE is a simple environment, represented using a matrix, where the agent's job is to pickup and deliver a package from a certain point in the matrix to another point specified earlier in the least number of steps possible
Image from Learn python with Rune
2 . Path Finder (in development) code
Path Finder is a simple environment, defined by and m X n square -- where m = no columns, n = no rows -- having obstacles at some fixed positions. The agent and destination are completely mobile on the board, i.e. they could be anywhere on the board except for the points where the obstacles are at.