make_diet_rand {qfasar}

Generate random diet compositions

Description

The function make_diet_rand generates a user-specified number of random diet compositions.

Usage

make_diet_rand(uniq_types, n_diet)

Arguments

uniq_types

A factor of unique prey-type names.

n_diet

The integer number of diet compositions to generate.

Value

A list containing the following elements:

diet_rand

A numeric matrix of random diet compositions, in column-major format.

err_code

An integer error code (0 if no error is detected).

err_message

A string contains a brief summary of the execution.

Details

The function make_diet_rand generates a specified number of random diet compositions to support simulation-based research of the performance of QFASA diet estimation procedures. Given a diet composition, predator fatty acid signatures can be generated using the function make_pred_sigs. The diets of such simulated predators can then be estimated, and the diet estimates can be compared to the known diet composition to evaluate bias, variance, and perhaps other properties.

The algorithm starts by generating a uniformly distributed random number between 0 and 1 as the diet proportion for the first prey type. The algorithm then considers each additional prey type in turn, generating a uniform random number between zero and 1 minus the sum of the proportions assigned to the preceding prey types. The diet proportion for the last prey type is 1 minus the sum of the other diet proportions. As a hedge against limitations in the random number generator, the proportions are then randomly ordered among prey types.

It is critical that the prey-type names match those in the prey library. The easiest way to ensure this happens is to pass the object uniq_types returned a call to the function prep_sig as the uniq_types argument. Alternatively, and more risky, a vector of unique prey names can be created using the concatenate function and cast as a factor, i.e., uniq_types <- as.factor(c("Prey_1", "Prey_2", ..., "Prey_P)).

Examples

make_diet_rand(uniq_types = as.factor(c("Bearded",
                                        "Beluga",
                                        "Bowhead",
                                        "Ribbon",
                                        "Ringed",
                                        "Spotted",
                                        "Walrus")),
               n_diet = 100)