Skip to contents

Maximum likelihood estimation of the Fisher parameters.

Source: R/inference.R
fisher-mle.Rd

MLE parameters describing a von Mises-Fisher distribution for isotropic directional vectors. Based on Mardia and Jupp (2000, p. 198).

Usage

fisher_MLE(x)

# S3 method for class 'Vec3'
fisher_MLE(x)

# S3 method for class 'Ray'
fisher_MLE(x)

# S3 method for class 'Plane'
fisher_MLE(x)

Arguments

x

object of class "Vec3", "Ray", or "Plane", where the rows are the observations and the columns are the coordinates.

Value

A list with members

muHat

the mean ray, identical to sph_mean()

rBar

non-negative real number. Mean resultant length.

kappaHat

a positive real number. Concentration parameter \(\kappa\) of the Fisher distribution

See also

fisher_inference() for confidence regions, and rvmf() to simulate a distribution. vmf_MLE() is an alternative MLE function.

Other distribution-MLE: bingham-mle, dist.mle, watson-mle

Examples

set.seed(20250411)
x <- rvmf(100, mu = Ray(120, 50), k = 5)

fisher_MLE(x)
#> $muHat
#> Ray object (n = 1):
#>   azimuth    plunge 
#> 118.32579  50.88713 
#> 
#> $rBar
#> [1] 0.8000961
#> 
#> $kappaHat
#> [1] 5.000538
#>