paired_SI.test: Paired Size Index (LSI, VSI, VSI*) test

In sginot/MorphoInd: Simpsons LSI and Uerpmanns VSI

Paired Size Index (LSI, VSI, VSI*) test

Description

This function will test for the difference in SI between sample(s) and the reference sample of the dataset. This tests descriptors individually, i.e. each morphological descriptor will have its own test.

Usage

paired_SI.test(dat, x, ref = 1, compsp = 1:length(dat))

Arguments

dat	Input data should be a list of matrices/data frames (same as for morpho_indices function).
x	Output from the morpho_indices function, containing the dataframes with SI values, as well as bootstraped values.
ref	Index of the reference sample (numeric or character), default is 1, so the first sample in 'dat' is taken to be the reference. Must be the same as defined for the morpho_indices function.
compsp	Vector of index(es) of the samples for which the difference is to be tested (numeric or character). Default is 1:length(dat), i.e. all samples, including the reference.

Details

NB: This test should be used only when the global test (global_SI.test function) showed significant global differences between sample(s) and reference. The test is based on H0: no significant difference between sample and reference. The statistic follows a Welch's T distribution. The expected input is 1) the original dataset, as formatted for the morpho_indices function (argument dat) and 2) the output from the morpho_indices function (argument x).

Value

mat.res A matrix containing observed values of the statistic, 0.95 quantile values for the statistic under H0 (i.e. critical statistic values for significance with alpha=0.05), and P-values (i.e. P(SI >= SI_obs)).

Examples

data("equus", package="MorphoInd")
o <- morpho_indices(dat=equus, ref=1, data.type="summary", bootstrap="p", plot=F)
test <- paired_SI.test(dat=equus, x=o, ref=1, compsp=2:10)

sginot/MorphoInd documentation built on Jan. 27, 2024, 1:32 a.m.