Rev Language Reference

dnPhyloCTMCDASiteIID(Tree tree, RateGenerator Q, Simplex rootFrequencies, RealPos branchRates, RealPos[] siteRates, Probability pInv, Natural nSites, String type, Bool treatAmbiguousAsGap, String coding)

tree :	Tree (pass by const reference)
	The tree along which the process evolves.
Q :	RateGenerator (pass by const reference)
	The global or branch-specific rate matrices.
rootFrequencies :	Simplex (pass by const reference)
	The root specific frequencies of the characters, if applicable.
	Default : NULL
branchRates :	RealPos (pass by const reference)
	The global or branch-specific rate multipliers.
	Default : 1
siteRates :	RealPos[] (pass by const reference)
	The rate categories for the sites.
	Default : [ ]
pInv :	Probability (pass by const reference)
	The probability of a site being invariant.
	Default : 0
nSites :	Natural (pass by value)
	The number of sites, used for simulation.
	Default : 10
type :	String (pass by value)
	The data type, used for simulation and initialization.
	Default : DNA
	Options : DNA|RNA|AA|PoMo|Protein|Standard|NaturalNumbers|Restriction
treatAmbiguousAsGap :	Bool (pass by value)
	Should we treat ambiguous characters as gaps/missing?
	Default : FALSE
coding :	String (pass by value)
	Default : all

AbstractHomologousDiscreteCharacterData

The parameters of a phylogenetic model -- a tree topology with branch lengths, a substitution model that describes how observations evolve over the tree, etc. -- collectively form a distribution called the _phylogenetic continuous-time Markov chain_. In this distribution, the branch history is sampled (through data augmentation), instead of integrated out (as in `dnPhyloCTMC`). An instance of character history can therefore be retrieved from the distribution (using `.characterHistories`). The probability of observed character state vectors (specified via clamping the distribution to an `AbstractHomologousDiscreteCharacterData` object) is computed by summing the probability of the history on each branch. The probability of the history on a branch is the product of the probabilities of waiting times between events (or the probability of no event in the final segment) given the current rate of change (see May and Moore 2020). Note that when `rootFrequencies` is not provided, the distribution assumes stationary frequencies at the root. The stationary frequencies will be calculated numerically if the tree has a root branch, and analytically otherwise. When `rootFrequencies` is provided, then stationarity at the root will not be assumed, and the likelihood calculation will be based on the provided root frequencies. If this distribution is used together with `dnPhyloOUSD` or `dnPhyloBMSD`, specify `nSites=1`.

# Read character data from a file
chars <- readDiscreteCharacterData("myData.nex")
taxa = chars.taxa()

# Draw a tree with branch lengths
tree ~ dnUniformTopologyBranchLength( taxa, branchLengthDistribution=dnExp(10.0) )

# Define a rate matrix
q_matrix <- fnJC(4)

rf <- rep(1/4, 4)

# Create stochastic node with the tip distribution given by `tree` and `q_matrix`
x ~ dnPhyloCTMCDASiteIID(tree = tree, Q = q_matrix, rootfrequencies=rf)
char_hist := x.characterHistories()

# Clamp observed characters to the node
x.clamp(chars)

# Calculate the probability of the observed characters under the given distribution
x.lnProbability()

• mvCharacterHistory