pseudo.pred: Generate a pseudo predator by sampling with replacement from...
In QFASA: Quantitative Fatty Acid Signature Analysis

Description Usage Arguments Details Value Examples

Generates a single pseudo predator by sampling with replacement from prey database. To generate a sample of pseudo predators, please refer to example code.

1	pseudo.pred(diet, preybase, cal.vec, fat.vec, preysize = 2)

`diet`	the "true" of "desired" diet of the pseudo predator. A compositional vector of proportions that sum to one with length equal to the number of prey species.
`preybase`	prey database from which to generate the pseudo predator. First column must provide the species name.
`cal.vec`	vector of calibration coefficients whose length is the same as the number of fatty acids in prey database.
`fat.vec`	vector of fat content whose length is the same as the number of species.
`preysize`	number of prey to sample from prey database. If preysize=1, then one prey is selected from each species. Otherwise, a sample of n_k signatures (where n_k is sample size for species k) is obtained by sampling with replacement.

The default is to re-sample all of the prey signatures within each species (that is, preysize=2). Alternatively, one prey may be randomly selected from each species yielding potentially more variable pseudo-predators. For details on simulating realistic predators signatures, see Bromaghin, J. (2015) Simulating realistic predator signatures in quantitative fatty acid signature analysis, Ecological Informatics, 30, 68-71.

A simulated predator FA signature.

data(preyFAs)

# Generating a sample of 10 pseudo predators each with "true" diet being
# (1/11,1/11,...1/11), no calibration effect and no fat content.  The QFASA diet estimate
# is then computed for each pseudo predator.

# Note: To incorporate calibration and fat content in a simulation study,
# one set of calibration and fat content is generally used to simulate the pseudo predator
# and another is used to estimate the diet.

set.seed(11)
p.mat <- matrix(rep(NA,10*11),nrow=10)
for (i in 1: 10) {
    my.seal <- pseudo.pred(rep(1/11,11),
                            preyFAs[,-c(1,3)],
                            rep(1,ncol(preyFAs[,-c(1,3)])-1),
                            rep(1,11))
     p.mat[i,] <- p.QFASA(my.seal,
                          MEANmeth(preyFAs[,-c(1,3)]),
                          rep(1,length(my.seal)),
                          2,
                          ext.fa=colnames(preyFAs[,-c(1:3)]))$`Diet Estimates`
 }

# Can verify that average diet estimate of the 10 pseudo predators is close to
# "true" diet.

round(apply(p.mat,2,mean),3)