The compr package provides functionality to fit and interpret paired competition models in R.

To install and load the package, run

devtools::install_github(repo = "https://github.com/jameswolpe/compr")

library(compr)

The compr package is designed to facilitate the fitting and interpretation of paired competition models in R. The primary function, comp_glm(), is the core feature of this package. It allows users to model paired competition outcomes using a formula that includes the response variable and contest-level predictors.

To install and load the compr package, use the following commands:

devtools::install_github(repo = "https://github.com/jameswolpe/compr")
library(compr)

The main function, comp_glm(), has the following arguments:

formula: Specifies the response variable and contest-level predictors. data: The dataset to be used. p1: A string with the name of the column containing the first player. p2: A string with the name of the column containing the second player. p1_effects: A one-sided formula specifying player-level effects for p1. p2_effects: A one-sided formula specifying player-level effects for p2. ref_player: A string giving the name of the player to be used as the reference player (default is the first player in p1).

Here is an example of using comp_glm():

comp_mod <- comp_glm(point_winner ~ -1, data = tennis_point,
                     p1 = "player1", p2 = "player2",
                     p1_effects = ~ point_server1,
                     p2_effects = ~ point_server2,
                     ref_player = "Milos Raonic")

Feel free to explore and leverage the compr package to enhance your paired competition modeling in R. For detailed information and usage examples, refer to the package documentation.