Maximum Likelihood Estimation of the Dirichlet Distribution — dirichlet.mle • sirt

Maximum likelihood estimation of the parameters of the Dirichlet distribution

Usage

dirichlet.mle(x, weights=NULL, eps=10^(-5), convcrit=1e-05, maxit=1000,
     oldfac=.3, progress=FALSE)

Arguments

x: Data frame with \(N\) observations and \(K\) variables of a Dirichlet distribution
weights: Optional vector of frequency weights
eps: Tolerance number which is added to prevent from logarithms of zero
convcrit: Convergence criterion
maxit: Maximum number of iterations
oldfac: Convergence acceleration factor. It must be a parameter between 0 and 1.
progress: Display iteration progress?

Value

A list with following entries

alpha: Vector of \(\alpha\) parameters
alpha0: The concentration parameter \(\alpha_0=\sum_k \alpha_k\)
xsi: Vector of proportions \(\xi_k=\alpha_k / \alpha_0\)

References

Minka, T. P. (2012). Estimating a Dirichlet distribution. Technical Report.

See also

For simulating Dirichlet vectors with matrix-wise \(\bold{\alpha}\) parameters see dirichlet.simul.

For a variety of functions concerning the Dirichlet distribution see the DirichletReg package.

Examples