encounterDistribution: Spatial distribution of encounters
In moveBB: Analysis of Trajectory Data using Linear or Brownian Motion Model

Description Usage Arguments Details Value Warning Note Examples

Computes the expected duration of encounters at each location for every pair of IDs.

1 2	encounterDistribution(tr, threshold, grid=NULL, timestepSize = 60, xc=NULL, yc=NULL, grid.dim = 100, grid.pad = 0.2)

`tr`	The trajectory for which to compute the UD
`threshold`	The maximum distance at which an encounter occurs
`grid`	a grid of class `asc`
`timestepSize`	The difference between consecutive time steps, in seconds
`xc`	The x coordinates of the vertical grid lines
`yc`	The y coordinates of the horizontal grid lines
`grid.dim`	If all of `grid`, `xc` and `yc` are omitted, this specifies the dimension of the shortest axis of the grid used.
`grid.pad`	If the grid is automatically generated, its range is the range of the relocations extended `grid.pad` times the length of the range on each side.

One can specify the grid in three ways:

If grid is set, the coordinates of the grid lines are derived from there.
These coordinates can also be specified in the parameters xc and yc.
Otherwise, the grid is determined automatically. The grid then ranges over a bounding box of all measurements, extended on each side by a specified fraction of the range. The number of grid cells can be controlled via grid.dim. The amount by which the grid is extended is controlled via grid.pad, which may be a vector specifying the extension on top,right,bottom,left respectively. It is recycled as usual.

The return value is a list, indexed by two IDs. If grid is given, each element of the result list is an object of class asc, representing the same grid. Otherwise, each element of the list is a matrix, indexed by the coordinates specified in xc and yc.

Element result[["id1", "id2"]] of the result represents the distribution of the position of id1 while it had encounters with id2. This is not the same as result[["id2","id1"]], since that is the distribution of id2's location during its encounters with id1.

The diagonal entries of the result list contain the utilization distribution of each ID, since an entity is always at a distance 0 from itself.

There seems to be some problem with the result being transposed, this needs further investigation. Until then, you can plot the transpose of the result using image(t(ud[["BD","NH"]])).

The image function has ugly colours, use the col attribute to define a better colour map.

Also note that this function may take a rather long time to complete, so please be patient, specify a sufficiently small grid or use a larger time step.

data("vervet_monkeys", package="moveBB")

## Define grid lines: equally spaced between the min and max coordinate in monkey.tr
#TODO: define from monkey.tr@extent
xlim <- range(monkey.data$X, na.rm=TRUE)
xc <- seq(xlim[1], xlim[2], length.out=20)

ylim <- range(monkey.data$Y, na.rm=TRUE)
yc <- seq(ylim[1], ylim[2], length.out=20)

## Compute the UD and plot the result for one ID
#ud <- encounterDistribution(monkey.tr, 100, xc=xc, yc=yc)
#image(ud[["BD","NH"]])