Spent much of the day trying to code an intelligent data-preprocessing function. Seems to mostly be working now, see github for details. Testing on a variety of datasets.
Currently the function assumes paintings can be specified by common ancestors of a set of tips, assumes non-overlapping clades. Doesn’t error-check for this either.
should also have more output processing functions, particularly to return a “phylo” object together with paintings (see earlier existing code in my Comparative-Phylogenetics repository).
Exploration on new data sets
- Carnivora – running with aquatic/terrestrial encoding and marine/freshwater/terrestrial encoding. Both set to run overnight. Wrightscape’s simplex method has some trouble with certain initial conditions on the three-regime model.
- Mammalia – body size, with regime by biome (aquatic, terrestrial, volant (bats). Huge data set, also running over night.
- Cetacea – body size by diet. my preprocessing function seems to have some trouble with the painting