dummy-link

DirectionalRegression

Perform regression using a directionally valued dependent variable.

Readme

Build Status codecov

We can fit a Von Mises model like this

using Distributions, DirectionalRegression
X = 2*rand(1000,3)-1
β = [0.181689,0.533594,0.132475]
μ0 = g(X*β)
θ = zeros(size(X,1))
for i in 1:length(θ)
     θ[i] = rand(sampler(VonMises(μ + μ0[i], κ)))
end
rr = DirectionalRegression.fitmodel(X, θ, 2*pi*rand()-pi, 0.1+3*rand(); tol=1e-3, maxiter=100)

First Commit

05/10/2016

Last Touched

over 4 years ago

Commits

15 commits

Used By: