Memory.Random.Play.PD: Generate Memory where strategies play against a random...

In MartinKies/RLR: Reinforcement Learning with R

Description

Each strategy within the game.object plays against a random strategy of the given defection probability

Usage

Memory.Random.Play.PD(game.object, algo.par)

Arguments

game.object	as specified by Get.Game.Object
algo.par	as e.g. given by Get.Def.Par.QLearning

Details

Outputs List of lists with the following elements:

• state - Already encoded game state, if algo.par$mem.type=="game.encoded"

• action - Which of the actions has been taken?

• next.state - resulting next, encoded, state

• reward - What did we get from transitioning to the next state?

• done - Boolean; is the game over?

Public Function which might be called by algorithms.

MartinKies/RLR documentation built on Dec. 24, 2019, 10:02 p.m.