IC: Compute information criterion scores and Akaike weights for evoltuionary models
Description
These functions compute information criteria (IC) or Akaike weights based on information scores (akaike.wts).
Function IC is used internally by the optimization functions and generally will not need to be called directly by the user.
Usage
IC(w, logL = NULL, K = NULL, n = NULL, meth = c("AICc", "AIC", "BIC"))
akaike.wts(aa)
Arguments
w
output from an optimization such as opt.GRW
logL
log-likelihood
K
the number of free parameters
n
sample size for AICc and BIC calculations
meth
which information criterion to compute; one of AIC, AICc, or BIC
aa
vector of AIC or AICc values used to compute Akaike weights
Value
the computed information criterion, or a vector of Akaike weights
Details
Function IC can take two kinds of arguments. The easiest is to send it the output of any of the paleoTS
optimizations, in which case the function will automaically extract the necessary information. Alternatively, the
log-likelihoods, number of parameters and sample size can be passed explicitly.
Function akaike.wts takes a vector of AIC or AICc values and computes corresponding Akaike weights.
References
Hunt, G. 2006. Fitting and comparing models of phyletic evolution: random walks and beyond. Paleobiology32:578--601.
x <- sim.GRW(ns=40, ms=0.1, vs=0.1)
m1<- opt.GRW(x)
m2<- opt.URW(x)
m3<- opt.Stasis(x)
akaike.wts(c(m1$AICc, m2$AICc, m3$AICc)) # it is easier to use fit3models()