Rev Language Reference

fnMinBLTimeScaling(Tree unscaledTree, Taxon[] taxa, RealPos minBL)

unscaledTree :	Tree (pass by const reference)
	The tree to be scaled.
taxa :	Taxon[] (pass by const reference)
	The vector of taxa; has to match the tips of the tree.
minBL :	RealPos (pass by const reference)
	Minimum branch length to use for time scaling.

TimeTree

The age of each internal node is based on the age of the oldest tip descended from it. However, if t0 is the oldest tip descended from a given node (denoted x) and also the oldest tip descended from that node's parent (denoted y), then setting t(x) = t(y) = t(t0) would produce zero-length branches y->x and x->t0. We avoid this by requiring every branch to be no shorter than some user-supplied constant. This has the effect of shifting node ages deeper into the past. Conceptually, the undated tree would usually correspond either to a bare topology (a tree without branch lengths) or a tree with branch lengths in units of expected change; in practice, both `BranchLengthTree` and `TimeTree` arguments are accepted. In this implementation of the MBL approach, both terminal and internal branches are required to be greater than or equal to the specified minimum. If there is uncertainty associated with the age of a given tip, the midpoint of the uncertainty range is used for time-scaling. The algorithm is not stochastic (i.e., it always returns the same time-scaled tree for a given input), and is primarily intended to generate a plausible starting tree for MCMC analyses.

# Read in an undated tree
undated_tree <- readTrees("undated.nex")[1]

# Read tip age data from a file
taxa <- readTaxonData("tipages.tsv")

# Time-scale using a minimum branch length of 3 Myr
dated_tree <- fnMinBLTimeScaling(undated_tree, taxa, 3.0)

print(undated_tree) # The original tree remains unchanged
print(dated_tree)   # A new, dated tree has been returned

• simStartingTree