Twelve volumes and 102 days of forty-five minutes. Every line of code is copyable with its real output printed underneath, so nothing depends on your machine working first. The course is built backwards from what a PhD committee actually checks: that you chose the right model, that you can say why, and that someone else can rerun you.
Volumes one to three are craft; four to eight are the statistics your research actually runs on; nine turns the work into evidence. Each keeps its own progress in this browser.
Every volume is the same shape, so once you have done day 1 you know the whole course. Forty-five minutes, one sitting, in this order.
library() this session; ?function_name in the console; then the error decoder in Foundations. Ninety per cent of first-month errors are a typo, an unloaded package, or a column that is a character when you thought it was a number.Most people learn commands and still pick the wrong test. Answer three questions about your study and this gives you the model, the call, the effect size to report, and the mistake reviewers look for first.
Type a function, a package, or what you are trying to do. Results say which day teaches it.
?function_name in the console is the canonical answer.Courses produce certificates; applications need evidence. Each of these is produced as a by-product of a specific day, not as extra work. Tick them off as they exist.
Arithmetic on the real minute estimates in these volumes, not an average. Pick the days per week you will actually manage.
Package sprawl is how beginners lose a week. Run these three lines on day 1; every later volume names its own additions on the day you need them.
Installing happens once per machine. library() happens once per session. Confusing the two is the most common first-week error, and the error message never says so.