auto.a: Parameters to select a reasonable value of 'a'

In tlocoh: Constructs homeranges and explores time use patterns from location data

Description

Returns a list of parameters that other functions accept for the 'a' parameter

Usage

auto.a(meth = "nn", ptp = 0.98, nnn = 2, tct = 1.05)

Arguments

meth	The name of the method for finding a (default is "nn" method for 'nearest neighbor' method.
ptp	The proportion of total points [0..1]
nnn	The number of nearest neighbors
tct	The temporal continuity threshhold, see details.

Details

If multiple values are passed for any of the parameters, the resulting data frame will have multiple rows and other parameters will be recycled as needed.

This function does not actually find nearest neighgors. Rather it returns a list of parameters that can be used as input in other functions such as lxy.nn.add.

Value

A data frame with columns meth, ptp, nnn, and tct which are the settings that other functions will use to find a value for 'a' should be automatically selected. Columns include:

• meth The name of the method for finding a

• ptp The proportion of parent points that should have at least a.nn neighbors (default 0.98)

• nnn The minimum number of nearest neighbors that each point will get (default 2, so each point can be part of a triangle)

• tct A time threshhold (expressed as a percentage of the median sampling frequency), such that if a point is temporally disconnected from its 2 nearest temporal neighbors by more than than tct times the median sampling frequency, the point will be be excluded from the computation of a. This is to avoid bias from internal time gaps (default 1.05)

See Also

lxy.nn.add

Examples

## Not run: 
lxy <- lxy.nn.add(lxy, ptsh=0.5, a=auto.a())

## End(Not run)

tlocoh documentation built on May 2, 2019, 5:27 p.m.