p.MLE {QFASA}

Returns simplified MLE diet estimates corresponding to a sample of predators.

Description

Computes the diet estimate for each predator in pred.mat using the simplified MLE method, without the use of random effects.

Usage

p.MLE(pred.mat, prey.mat, cal.mat, FC, ext.fa)

Arguments

pred.mat

matrix containing the FA signatures of the predators.

prey.mat

matrix containing the FA signatures of the individual prey. The first column must index the prey group. prey.mat is the prey database.

cal.mat

matrix of calibration factors where the i th column is to be used with the i th predator. If modelling is to be done without calibration coefficients, simply pass a vector or matrix of ones.

FC

vector of fat content of length equal to the number of prey groups or species.

ext.fa

subset of fatty acids to be used to obtain diet estimates.

Details

The assumed model is similar to the MUFASA model but the random effects are replaced by the prey species' sample means to speed up computations. Unlike p.MUFASA, this function does not require integration and hence is much faster.

Value

A list with components:

Diet_Estimates

A matrix of the diet estimates for each predator where each row corresponds to a predator and the columns to prey species. The estimates are expressed as proportions summing to one.

Var_Epsilon

Optimized values of error variance. See reference.

nll

Negative log likelihood values. As per solnp documentation, nll is a "vector of function values during optimization with last one the value at the optimal".

References

Steeves, Holly (2020) Maximum likelihood approach to diet estimation and inference based on fatty acid signatures. PhD thesis available at https://dalspace.library.dal.ca/handle/10222/80034.

Examples

##  This example takes some time to run.
## Please uncomment the code below to run.


#library(dplyr)
#library(compositions)


## Fatty Acids
#data(FAset)
#ext.fa <- as.vector(unlist(FAset))

## Predators
#data(predatorFAs)
#pred.mat <- predatorFAs[, -c(1:4)]
#n.pred <- nrow(pred.mat)

## Prey
#data(preyFAs)
#prey.mat <- preyFAs[, -c(1,3)]

#FC = preyFAs[,c(2,3)]
#FC = as.vector(tapply(FC$lipid,FC$Species,mean,na.rm=TRUE))

## Calibration Coefficients
#data(CC)
#cal.vec = CC[,2]
#cal.mat = replicate(n.pred, cal.vec)
#rownames(cal.mat) <- CC$FA


## Diet Estimates
#mle.est <- p.MLE(pred.mat, prey.mat, cal.mat, FC, ext.fa)
#mle.est$"Diet Estimates"