Directional (version 4.0)

Contour plot of spherical data using a von Mises-Fisher kernel density estimate: Contour plot of spherical data using a von Mises-Fisher kernel density estimate

Description

Contour plot of spherical data using a von Mises-Fisher kernel density estimate.

Usage

vmf.kerncontour(u, thumb = "none", den.ret = FALSE, full = FALSE, ngrid = 100)

Arguments

u

A two column matrix. The first coolumn is the latitude and the second is the longitude.

thumb

This is either 'none' (defualt), or 'rot' for the rule of thumb suggested by Garcia-Portugues (2013). If it is "none" it is estimated via cross validation, with the fast function "vmfkde.tune_2".

den.ret

If FALSE (default), plots the contours of the density along with the individual points. If TRUE, will instead return a list with the Longitudes, Latitudes and Densities. Look at the 'value' section for details.

full

If FALSE (default), uses the range of positions from 'u' to calculate and optionally plot densities. If TRUE, calculates densities covering the entire sphere.

ngrid

Sets the resolution of the density calculation.

Value

The contour lines of the data. If "den.ret" was set to TRUE a list including:

lat

The latitude values.

long

The longitude values.

h

The optimal bandwidth.

den

The kernel density estimate contour points.

Details

It calcculates the contour plot using a von Mises-Fisher kernel for spherical data only.

References

Garcia Portugues, E. (2013). Exact risk improvement of bandwidth selectors for kernel density estimation with directional data. Electronic Journal of Statistics, 7, 1655--1685.

See Also

vmf.kde, vmfkde.tune, vmf.contour, kent.datacontour

Examples

# NOT RUN {
x <- rvmf(100, rnorm(3), 15)
x <- euclid.inv(x)
par( mfrow = c(1, 2) )
vmf.kerncontour(x, "rot")
vmf.kerncontour(x, "none")
# }