loglik: Calculates the log likelihood of a population model, given a...
In UCL/ADMUR: Ancient Demographic Modelling Using Radiocarbon

loglik

R Documentation

Calculates the log likelihood of a population model, given a calibrated date PD matrix

Description

Calculates the log likelihood of a population model, given a calibrated date PD matrix

Usage

loglik(PD, model)

Arguments

`PD`	A data frame of Probability Distributions (PDs). Each column representing the PD of a calibrated observation or phase. Row names correspond to the calendar years BP. This data frame can be generated by `phaseCalibrator`
`model`	A data frame containing the columns 'pdf' representing a hypothesised population Probability Density Function; and 'year' corresponding to the calendar years BP.

Details

Row names of both PD and model arguments must exactly match, since the probability of each phase given the model is calculated numerically.

Value

Returns a single numeric log likelihood.

Examples

	# calibrate a dataset comprising just two phases
	data <- data.frame(age=c(5800, 5100),sd=c(40, 35),phase=c('p1', 'p2'), datingType='14C')
	CalArray <- makeCalArray(shcal20, calrange = range(toy$year))
	PD <- phaseCalibrator(data, CalArray)	

	# calculate toy model log likelihood
	loglik(PD, toy)

UCL/ADMUR documentation built on Sept. 14, 2023, 11:41 a.m.