Maximum likelihood estimation of Spherical Rotational Symmetric Distributions

MLE parameters of a von Mises-Fisher or Kent distribution.

Usage

kent_MLE(x)

vmf_MLE(x)

Source

Adapted from Directional::kent.mle() and Directional::vmf.mle()

Arguments

x

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

See also

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

Other distribution-MLE: bingham-mle, fisher-mle, watson-mle

Examples

set.seed(20250411)
x <- rvmf(100, mu = Ray(120, 50), k = 5)
vmf_MLE(x)
#> $loglik
#> [1] -122.7913
#> 
#> $mu
#> Ray object (n = 1):
#>   azimuth    plunge 
#> 118.32579  50.88713 
#> 
#> $kappa
#> [1] 5.000134
#> 
 
x2 <- rkent(100, mu = Line(120, 50), k = 5, b = 1)
kent_MLE(x2)
#> $G
#> Ray object (n = 3):
#>       azimuth    plunge
#> [1,] 122.8747 50.591038
#> [2,] 210.0219 -2.341775
#> [3,] 298.1030 39.311419
#> 
#> $param
#>      kappa       beta        psi 
#>  4.4706190 -0.6586224  0.3987251 
#> 
#> $logcon
#> [1] 4.835128
#> 
#> $loglik
#> [1] -134.7842
#>