pctn: Principal coordinates of temporal neighbours

In gavinsimpson/temporalEF: Temporal Eigenfunctions

Description

Computes the classic PCNM by the principal coordinate analysis of a truncated distance matrix, but for a one-dimensional process.

Usage

pctn(x, ...)

## Default S3 method:
pctn(x, threshold, distfun = dist, ...)

## S3 method for class 'pctn'
print(x, digits = 3, ...)

## S3 method for class 'pctn'
scores(x, choices, ...)

## S3 method for class 'pctn'
eigenvals(x, ...)

Arguments

x	an R object. For pctn currently only a sorted vector of time points.
...	additional arguments passed to other methods or on to distfun.
threshold	numeric; threshold beyond which the temporal separation of smaples is considered equal. The default if no value is supplied is to find the largest temporal separation between any two points. Separations greater than the threshold are given a notional separation of 4 times threshold.
distfun	function or character string naming a function that will be used to compute the temporal seperation between samples. Defaults to dist for the Euclidean distance. See Details for further information.
digits	numeric; number of digits to display in output.
choices	numeric; vector indicating which eigenfunctions to return.

Details

The default distance coefficient used to compute temporal separation is the Euclidean distance. If you want to use a different coefficient, you can supply a suitable function to argument distfun. This should be a function that returns an object of class "dist" or a square symmetric matrix that can be coerced to one. Arguments can be passed to distfun via ....

Author(s)

Gavin L. Simpson

Examples

tp <- seq_len(50)
mod <- pctn(tp)
mod

gavinsimpson/temporalEF documentation built on May 16, 2019, 10:11 p.m.