getLogFit: Get the Logistic Regression Fit

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

Description

This function takes in a training data set and performs a logistic regression model to predict the stay probability as a function of whether the last trial was rewarded or not, whether the last trial had a common or rare transition, and the interaction between the two. This can be run by itself, but is called by the function getLogPreds.

Usage

getLogFit(data)

Arguments

data	The data can either be the output of logSetup or the output of generateData. If the latter, make sure that manipulateData is called on pure Q-learning simulations. Similarly, for DynaQ simulations, make sure the data has been processed and the Q table has been removed.

Value

This returns a tidymodels workflow object that has been fitted to predict the stay probability based on the reward and transition types.

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