preACF: Estimate the autocorrelation function before interpolating...

In lawlerem/markmodmover: Analyze animal movement data using the Markov-modulated movement model.

Description

Estimate the acf for step length before interpolating the data.

Usage

preACF(object, ...)

## S4 method for signature 'Data4M'
preACF(object, Time.Step, units = "hrs",
  lag.max = 30, kernel = "gaussian", kernel.width = "default")

Arguments

object	An object of or inheriting from class Data4M.
Time.Step	The time interval used for the regular time grid.
units	Can be one of days, hrs, mins, or secs.
lag.max	The maximum lag to compute.
kernel.width	The bandwidth of the kernel used in weighting pairs of observations. Note that the time differences are standardized to have mean 1. The default setting depends on the kernel used.
kernel.type	Which kernel to use? Can be one of "rectangle", "sinc", or "gaussian". Can also supply a user-defined function that takes the arguments d, kernel.width, and n.

Value

An object of class acf, which is a list containing the following elements:

• lag: A vector containing the lags at which the acf is estimated.

• acf: An vector with the same length as 'lag' containing the estimated acf.

• type: Always set to "correlation".

• n.used A vector giving the sum of kernel values for each lag.

• series The name of the series 'object'.

• snames Always set to NULL.

lawlerem/markmodmover documentation built on Feb. 12, 2020, 8:30 p.m.