Making an R Package

Transcript

1. Making an R package
2. Who am I? - Nick Williams - 2nd year MPH Candidate, Biostatistics Department - Have been working in R for real work and as a hobby for about 5-years
3. How I got into making a package - Integrating R into the curriculum for Analysis of Categorical Data (P8120) - Many of the R functions for basic categorical data analysis were outdated, spread across multiple packages, not beginner friendly, and followed different syntax - My goal: Consolidate and repurpose existing functions into one package with a unified structure - The result was catfun (categorical functions, categorical is fun, cats are fun)
4. Who can/should make a package? Anyone with: - With a solution to a problem that others likely also have - The technical know-how - A healthy dose of curiosity - So you
5. Basic package structure At a minimum, your package should contain at least 3 items: - R/ - DESCRIPTION - NAMESPACE If you want to publish on CRAN - tests/ - man/
6. What you need to consider - Your audience - Your goals - OOP (maybe) - Memory/speed (Rcpp) - The time to keep it functional For a comprehensive walk through: - R packages by Hadley Wickham
7. Let’s make a package!