#Git for researchers

Rerun Your Analysis Two Years Later and Get the Same Numbers: A Workflow

A reviewer asks you to rerun one model and you cannot reproduce your own Table 3. Folder structure, code instead of clicks, recorded package versions and random