overlapPerm: Overlap window permutation tests.
In jbhender/camtrap: Inference for camera trap data.

Description Usage Arguments Details Value See Also

Assess the signficance of the overlap or excess in a window using a simulation based permutation test.

1 2	overlapPerm(A, B, t0 = 0, t1 = 24, adjust = 0.8, nperm = 1000, kmax = 3, parametric = T, two.sided = T, excessRef = 0)

`A`	Numeric vectors of sighting times in [0,2pi]
`B`	Numeric vectors of sighting times in [0,2pi]
`t0`	The window over which to estimate the overlap and excess stats
`t1`	The window over which to estimate the overlap and excess stats
`adjust`	See `overlapEst`.
`nperm`	The number of simulated samples or permutations to generate.
`kmax`	See `overlapEst`.
`two.sided`	If `TRUE` the p-values returned for the excess are two-sided. P-values for the overlap are always one-sided.
`excessRef`	The reference value against which to compare the observed excess. The default is zero, which should make sense in most cases.

This function estimates the probability of observing an overlap/excess in a given interval as extreme as the one observed under the null hypothesis that A and B actually come from the same distribution. By default, this hypothesis is tested by simulating new data A' and B' from a Null density $f_0$ estimated by concatenating A and B. Setting parametric = F will instead give a traditional permutation test (not yet implemented.)

An object of class overlapPermObj; a list containing

table: A table listing the observed overlap and excess in the window, their reference values under the null, and p-values from the permutation tests.
window: The requested window c(t0,t1)
nperm: The requested number of permtutations
nperm,parametric,two.sided,: The values passed in the funciton call.
overlapRef: The esitmated reference value, computed as: $ \int_t_0^t_1 \hat f_0(s) ds $