README.md

Quick start

Welcome to the codenames GitHub page!

This package culminates from an introductory exercise in game AI. As an avid fan of the board game Codenames by Vlaada Chvatil, I sought to develop a simple AI that would, at each state in the game, proffer a word that is maximally suggestive of one set of words while minimally suggestive of another set of words. To accomplish this, I used the Edinburgh Associative Thesaurus (EAT) in conjunction with a custom feature selection function to prioritize word clues from a database of several thousand words. You can get started with this implementation of codenames, complete with AI, by installing the most up-to-date version of this package directly from GitHub.

library(devtools)
devtools::install_github("tpq/codenames")
library(codenames)

Begin playing

This package includes everything needed to play a Codenames-like word association game from within the R console. To launch the game, users simply need to call the codenames function from this package. Players can select 0, 1, or 2 AI "spy masters" by toggling the function arguments sm1 and sm2 between "human" and cpu. I refer those unfamiliar with the rules to the excellent tutorial video here.

codenames(sm1 = "human", sm2 = "cpu")

By default, each executed game will continue to run until completed. To terminate the game early, enter kill or quit during a user prompt. To pass on a turn, enter pass during a user prompt. Otherwise, follow all directions provided within the console, and have fun!



tpq/codenames documentation built on May 31, 2019, 6:50 p.m.