MQMF (version 0.1.0)

negLL1: negLL1 calculate log-normal log-likelihoods with a penalty

Description

negLL1 calculates log-normal negative log-likelihoods. It expects the input parameters to be log-transformed, so the funk used to calculate the log or the predicted values also needs to expect log-transformed parameters. In addition to estimating the negative log-liklelihoods for log-normally distributed data it also places a penalty on the first parameter if that parameter approaches very close to zero; see the function penalty0. With SPM the first parameter is the population growth rate r, which obviously should never be negative. The use of penalty0 prevents this.

Usage

negLL1(pars, funk, logobs, ...)

Arguments

pars

the log-transformed parameters to be used in the funk for calculating the log of the predicted values against which the log observed values will be compared

funk

the function used to calculate the log-predicted values of whatever variable is being used (eg. cpue, catches, etc.)

logobs

the observed values log-transformed ready for comparison with the log-predicted values from funk and pars.

...

required to allow funk to access its other parameters without having to explicitly declare them in negLL

Value

the negative log-likelihood using log-normal errors.

Examples

# NOT RUN {
data(abdat)  #expect an answer of -31.65035
param <- log(c(r= 0.42,K=9400,Binit=3400,sigma=0.05))
negLL1(pars=param,funk=simpspm,logobs=log(abdat[,"cpue"]),indat=abdat)
# }