| Title: | Matrix Skew-T Parameter Estimation |
|---|---|
| Description: | Performs matrix skew-t parameter estimation, Gallaugher and McNicholas (2017) <doi: 10.1002/sta4.143>. |
| Authors: | Michael P.B. Gallaugher, Paul D. McNicholas |
| Maintainer: | Michael P.B. Gallaugher <[email protected]> |
| License: | GPL (>= 2) |
| Version: | 0.1.5 |
| Built: | 2024-12-18 06:49:03 UTC |
| Source: | CRAN |
Performs paramter estimation for the matrix variate skew-t distribution using an ECM algorithm.
Fit_Skewt(X, Tol = 0.001, max_iter = 1000)Fit_Skewt(X, Tol = 0.001, max_iter = 1000)
X |
A list of matrices of the same size |
Tol |
The tolerance of the ECM algorithm. Defaults to 0.001 |
max_iter |
The maximum number of iterations. Defaults to 1000 |
Returns a list with elements M (the estimate of the location), A (the estimate of the skewness), nu (the estimate of the degrees of freedom), Sigma (the estimate of Sigma), Psi (the estimate of Psi), loglik (a vector of log likelihood values), flag (returns TRUE if a numerical issue occured, FALSE otherwise).
data(SimX) Fit_st<-Fit_Skewt(SimX)data(SimX) Fit_st<-Fit_Skewt(SimX)
This is a simulated dataset with 100 observations from 4 by 3 matrix skew-t distribution.
data(SimX)data(SimX)
An object of class list of length 100.