Articles
Start Here
The tour: one analysis carried from the formula through to the estimates, with a pointer out of each section to the guide that takes it further.
Fitting a Model
Choosing the likelihood, which is the modeling decision that matters more than any sampler setting; the response type whose families need a guide of their own; and the two terms that add parameters to a sum of trees, a coefficient that varies and an intercept per group.
Reading a Fit
Turning a fitted forest into the quantities an analysis reports: effects and the shape of a relationship, which predictors the model spends its splitting rules on, and what it takes to read an effect as causal.
Checking a Fit
Whether the sampler converged and whether the model describes the data, and how to choose between two models that both do.
Reference
What the model is, how it is fitted, and why the defaults are what they are, with the short answers alongside.