Common Arguments of divent Functions

Functions to estimate entropy and diversity share many arguments, whose default value is generally the best choice. They are explained here.

Asymptotic or interpolated / extrapolated diversity

q

This argument sets the order of diversity to estimate, e.g. 0 for richness, 1 for Shannon’s and 2 for Simpson’s entropies.

estimator

Each function comes with several estimators that correspond to the state of the art. Asymptotic entropy and diversity are calculated according to the chosen estimator.

level

level allows estimating entropy or diversity at a chosen level, that may be a number of individuals (at least 1) of a sample coverage (between 0 and 1) that is first converted to a number of individuals by coverage_to_size().

If level is not NULL, estimator is ignored.

sample_coverage

To estimate \(\gamma\) diversity, the size of a metacommunity is unknown so it has to be set according to a rule which does not ensure that its abundances are integer values. Then, classical bias-correction methods do not apply. Providing the sample_coverage argument allows applying the ChaoShen and Grassberger estimators to estimate quite well the entropy.

Unveiling

The unveiled estimators, whose names start with unveil, rely on the estimation of the distribution of probabilities of the species (i.e. the probability for an individual to belong to each species). They require several arguments, that are ignored if another estimator is chosen.

probability_estimator

The probability estimators allow estimating the actual probability of species from the observed abundances.

unveiling

The unveiling technique sets the distribution of the unobserved species. If unveiling is “none”, they are ignored.

richness_estimator

The number of unobserved species is estimated by richness_estimator.

jack_alpha and jackmax

If richness_estimator is jackknife, jackmax is the maximum order allowed. jack_alpha is the risk level of the confidence interval of the estimation.

The estimation starts at order \(k=0\). If the estimate of order \(k+1\) is out of the confidence interval of the estimate of order \(k\), then \(k\) is incremented and the test is repeated until the optimal order is found.

coverage_estimator

Unveiling is based on the estimation of the sample coverage. Its estimator can be chosen.

General arguments

alpha

The risk level of tests or confidence intervals.

as_numeric

Diversity and entropy functions return a tibble with detailed information about the estimation. When applied to a single community, they can return a number instead of a one-line tibble with the argument as_numeric = TRUE.

check_arguments

Most functions of the package check their arguments, e.g. that alpha, the risk level, is between 0 and 1. To save time, check_arguments = FALSE skips the checks.

n_simulations

The number of simulations run to estimate confidence intervals.