Writing Good Software

Adapted from Software Carpentry

Overview

Questions:

  • How can I write software that other people can use?

Objectives:

  • Describe best practices for writing R and explain the justification for each

Keep Projects Organized

Keep your project folder structured, organized, and tidy. Create subfolders for:

  • code
  • data
  • documentation
  • results and figures

RStudio projects help, and you can automate setup with packages such as ProjectTemplate.

TipTip: ProjectTemplate 
install.packages("ProjectTemplate")
library("ProjectTemplate")
create.project("../my_project_2", merge.strategy = "allow.non.conflict")

Make Code Readable

Readable code matters more than clever code. Someone else should be able to understand your work quickly, and that someone is often you in six months.

Document What and Why

Comments should explain what problem you are solving and why you made a decision. The how is usually clear from the code itself.

Keep Code Modular

Separate your functions from analysis scripts and store them in a file you source() when you start a project. Small, single-purpose functions are easier to test and reuse.

Test Your Work

Make sure your functions do the right thing. Tests build confidence and prevent regressions.

Do Not Repeat Yourself

Repeated blocks of code are good candidates for functions. This makes your project easier to maintain and update.

Be Consistent in Style

Use consistent naming, indentation, and formatting. Style consistency makes collaboration easier.

Key Points

  • Keep your project folder structured and tidy
  • Document what and why, not just how
  • Break programs into short single-purpose functions
  • Write re-runnable tests
  • Do not repeat yourself
  • Be consistent in naming and style