ant: ANT diagnostic for permuted statistics

In SebastianSosa/ant: Animal Network Toolkit Software

ANT diagnostic for permuted statistics

Description

ANT method to make a diagnostic of all ANT permutation tests. This method adapts the diagnostic results according to the data input. The output is adapted to the type of test run. However, some outputs are common to all tests.

Usage

ant(x)

Arguments

x	an ANT object from functions: stat.t, stat.c, stat.lm, stat.glm, stat.glmm pr a numeric vector or a data frame with only numeric values.

Value

A list of two elements fot stat.cor, stat.t functions:

• A data frame with the permuted p-values (left side and right side), the confidence interval (25, 50 or 95) and the mean of the posterior distribution of the statistics of interest according to the statistical test (coefficient of correlation, t of students, or estimate(s))

• An histogram of the posterior distribution of the statistics of interest according to the statistical test.

#' A list of four elements fot stat.lm, stat.glm, stat.glmm functions:

• A data frame with the orginal stats of the model, and permuted p-values (left side and right side), the confidence interval (25, 50 or 95) and the mean of the posterior distribution of the statistics of interest according to the statistical test (coefficient of correlation, t of students, or estimate(s))

• Diagnostic plot of the original model

• An histogram of the posterior distribution of the statistics of interest according to the statistical test.

• a vector of the permutations that generates errors and for which new permutations were performed.

Author(s)

Sebastian Sosa, Ivan Puga-Gonzalez.

Examples

t=met.strength(sim.m,sim.df,1) # Computing network metric
t=perm.net.nl(t,labels='age',rf=NULL,nperm=10,progress=FALSE) # Node label permutations
r.c=stat.cor(t,'age','strength',progress=FALSE) # Permuted correlation test
r=ant(r.c)

SebastianSosa/ant documentation built on Sept. 23, 2023, 7:06 a.m.