cu_crw_covmat: Calculate correlation matrix for a set of times from a CRW...

In dsjohnson/crawlUtils: Enhance And Integrate the {crawl} Package For Spatial Analysis Of Telemetry Output

Calculate correlation matrix for a set of times from a CRW covariance function

Description

Using a correlation function created by cu_crw_covfun from a fitted CRW model a covariance (correlation) matrix is created for observations at the user provided times.

Usage

cu_crw_covmat(x, corr = TRUE, cf, E = 0)

Arguments

x	Either a crwFit object from a call to crwMLE, or a vector of times.
corr	Should the function return a correlation or covariance matrix? Defaults to corr = TRUE.
cf	Covariance function created from a crwFit object with a call to cu_crw_covfun.
E	The 'zero' time used for the covariance function. Defaults to E = 0. (See cu_crw_covfun).

Details

If x is a crwFit object, then the cf and E arguments are ignored. The resulting matrix is the covariance (correlation) matrix for the observed location times conditioned on the first location. Therefore, for n observed locations an (n-1) by (n-1) matrix will result. This is most useful for the effective sample size computation for a kernel density estimate.

Author(s)

Devin S. Johnson

dsjohnson/crawlUtils documentation built on Sept. 13, 2024, 1:34 p.m.