No packages match

binaryRL - Reinforcement Learning Tools for Two-Alternative Forced Choice Tasks

Tools for building Rescorla-Wagner Models for Two-Alternative Forced Choice tasks, commonly employed in psychological research. Most concepts and ideas within this R package are referenced from Sutton and Barto (2018) <ISBN:9780262039246>. The package allows for the intuitive definition of RL models using simple if-else statements and three basic models built into this R package are referenced from Niv et al. (2012) <doi:10.1523/JNEUROSCI.5498-10.2012>. Our approach to constructing and evaluating these computational models is informed by the guidelines proposed in Wilson & Collins (2019) <doi:10.7554/eLife.49547>. Example datasets included with the package are sourced from the work of Mason et al. (2024) <doi:10.3758/s13423-023-02415-x>.

Last updated

mapmlepsychologyrcppreinforcement-learningtafccpp

5.00 score 9 stars 6 scripts 597 downloads

multiRL - Reinforcement Learning Tools for Multi-Armed Bandit

A flexible general-purpose toolbox for implementing Rescorla-Wagner models in multi-armed bandit tasks. As the successor and functional extension of the 'binaryRL' package, 'multiRL' modularizes the Markov Decision Process (MDP) into six core components. This framework enables users to construct custom models via intuitive if-else syntax and define latent learning rules for agents. For parameter estimation, it provides both likelihood-based inference (MLE and MAP) and simulation-based inference (ABC and RNN), with full support for parallel processing across subjects. The workflow is highly standardized, featuring four main functions that strictly follow the four-step protocol (and ten rules) proposed by Wilson & Collins (2019) <doi:10.7554/eLife.49547>. Beyond the three built-in models (TD, RSTD, and Utility), users can easily derive new variants by declaring which variables are treated as free parameters.

Last updated

abcif-elsekeraskeras3mapmlemulti-armed-banditpcaplspsychologyrcppreinforcement-learningrnncpp

4.75 score 4 stars 2 scripts 340 downloads

abcpp - Shared C++ Backend for Approximate Bayesian Computation

Provides a compact C++ backend for Approximate Bayesian Computation (ABC) with a thin R frontend. The current implementation is primarily a C++ reimplementation of offline ABC workflows provided by the R 'abc' package <doi:10.32614/CRAN.package.abc>, with the public R interface intentionally kept small and centered on abc() and summary() methods. The computational work is performed by shared C++ code. In addition to reproducing common 'abc' workflows, the package adds optional dimensionality reduction of summary statistics through Principal Component Analysis (PCA) and Partial Least Squares (PLS), following related ideas described by Bazin et al. (2010) <doi:10.1534/genetics.109.112391> and Wegmann et al. (2009) <doi:10.1534/genetics.109.102509>.

Last updated

cpp

3.30 score 2 scripts