MQMF (version 0.1.0)

simpspm: simpspm calculates only the predicted log(CE) for an SPM

Description

simpspm calculates only the predicted log(CPUE) for a Surplus Production Model (SPM). It sets the Polacheck et al, 1993 parameter 'p' depending on the schaefer boolean argument, and this determines the asymmetry of the production curve. Note p is set not estimated. If p = 1.0 then the SPM is the Schaefer model, if it is 1e-8 it approximates the Fox model. The output of log(CPUE) is to simplify the use of log-normal residual errors or likelihoods. This function is designed for data consisting of only a single cpue time-series. simpspm must have at least three parameters, including the sigma, even if sum-of-squared residuals is used as a minimizer, then sigma would just float. The column titles for year, catch and cpue are included to facilitate ease of use with other data sets.

Usage

simpspm(
  pars,
  indat,
  schaefer = TRUE,
  year = "year",
  cats = "catch",
  index = "cpue"
)

Arguments

pars

the parameters of the SPM are either c(r,K,Binit,sigma), or c(r, K, sigma), the sigma is required in both cases. Binit is required if the fishery data starts after the stock has been depleted. Each parameter must be log-transformed for improved model stability and is transformed inside simpspm.

indat

the data which needs to include year, catch, and cpue.

schaefer

a logical value determining whether the spm is to be a simple Schaefer model (p=1) or approximately a Fox model (p=1e-08). The default is TRUE = Schaefer model

year

the column name within indat containing the years

cats

the column name within indat containing the catches

index

the column name within indat containing the cpue.

Value

a vector of length nyrs of log(cpue)

Examples

# NOT RUN {
 data(abdat)
 param <- log(c(r=0.4,K=9400,Binit=3400,sigma=0.05))
 predCE <- simpspm(pars=param,indat=abdat)
 cbind(abdat,exp(predCE))
# }