Learn R Programming

kairos (version 2.2.0)

assess: Statistical Significance of Seriation Solutions

Description

Tests the significance of seriation solutions.

Usage

assess(object, ...)

# S4 method for AveragePermutationOrder assess(object, axes = 1, n = 1000, progress = getOption("kairos.progress"))

Value

A list with the following elements:

random

A numeric vector giving the randomized total number of modes values.

observed

A numeric value giving the observed total number of modes.

expected

A numeric value giving the expected total number of modes if all types had unimodal distributions.

maximum

A numeric value giving the maximum total number of modes.

coef

A numeric value giving the seriation coefficient (a value close to 1 indicates a strong fit to the seriation model, while a value close to 0 indicates a poor fit).

Arguments

object

A PermutationOrder object giving the permutation order for rows and columns (typically returned by seriate_average()).

...

Currently not used.

axes

An integer vector giving the subscripts of the CA axes to be used.

n

A non-negative integer giving the number of bootstrap replications.

progress

A logical scalar: should a progress bar be displayed?

Author

N. Frerebeau

References

Porčić, M. (2013). The Goodness of Fit and Statistical Significance of Seriation Solutions. Journal of Archaeological Science, 40(12): 4552-4559. tools:::Rd_expr_doi("10.1016/j.jas.2013.07.013").

See Also

Other seriation methods: as_seriation(), order(), permute(), refine(), seriate_average(), seriate_rank()

Examples

Run this code
if (FALSE) {
## Data from Desachy 2004
data("compiegne", package = "folio")

## Correspondance analysis based seriation
(indices <- seriate_average(compiegne, margin = c(1, 2), axes = 1))

## Test significance of seriation results
## Warning: this may take a few seconds!
signif <- assess(indices, axes = 1, n = 1000)

## Histogram of randomized total number of modes
hist(signif$random)

## Observed value is smaller than the 5th percentile of the
## distribution of randomized samples
quantile(signif$random, probs = 0.05)
signif$observed

## Seriation coefficient
## (close to 1: relatively strong and significant signal of unimodality)
signif$coef
}

Run the code above in your browser using DataLab