manipulateData: Manipulate Data

In jdtrat/dynaq: Tools to Simulate DynaQ Reinforcement Learning Algorithms

Description

This function manipulates the output of generateData. It recodes actions as binary (right == 1, left == 0), calculates the proportion of actions right or left as of each trial, determines if a transition is common or not (inherent in the task structure), and filters the data to remove the Qtable while retaining meta data such as the temporal discounting factor (gamma) learning rate for simulated data (alpha), and the probability of receiving a reward for each image.

Usage

manipulateData(data)

Arguments

data	The data to be manipulated. Typically, the output of generateData, though it could be the output of oneTrial or simModel.

Value

Manipulated data with 21 columns and however many trials were run (just real or with real and simulated experience combined).

jdtrat/dynaq documentation built on July 24, 2020, 7:18 a.m.