In addition to the source code used for producing them, this release contains the simulation data used for the figures in The American Naturalist.
Files with .csv extension are in comma separated value format (without column header row). Files with .db extension are SQLite3 database files. Both have a common format, where the first column (named a6e in table br4nches in the .db files) is the full age of the simulated tree, and remaining columns g⟨ℓ⟩ are the counts of subbranches ℓ growth cycles younger than the branch being sampled; the averages of these values are thus what our function g(ℓ,n) should approximate. Files whose name begins with wholetree take the entire (simulated) tree as one sample, whereas those whose name begins with branches take each (again simulated) branch of a given age as a separate sample; that is why the latter are much larger.
The rest of the filenames encode the parameter values and growth method used in the simulation. For branches the order of parameters is µ-ß-d-p. For wholetree-rgrow the order is ß-d-p, whereas µ is random for each bud (at a distribution of 20%/60%/20% between values 1, 2, and 3); the same random µ distribution is used for wholetree-random, whereas it is µ=2 for wholetree-fixed. Remaining parameters for wholetree-fixed and wholetree-random are in order ß-d-p-q.