The **MVT** package contains a set of routines to perform estimation
and inference under the multivariate t-distribution. These methods are a direct
generalization of the multivariate inference under the gaussian assumption. In addition,
these procedures provide robust methods useful against outliers. The package uses these
functions in algorithms to simulate, analyze, and fit the model to data. The methodology
is described, with examples, in a working paper.

- Basic functionality for modeling using the multivariate t-distribution.
- Estimation of mean, covariance matrix and the shape (kurtosis) parameter using the EM algorithm.
- The core routines have been implemented in C and linked to R to ensure a reasonable computational speed.
- Performs several tests for testing the mean of a multivariate t population. Generalized Hotelling's T-squared, likelihood ratio test, score, Wald and gradient can be used as a test statistic.
- Performs hypothesis testing about the equicorrelation or homogeneity of variances structures for the covariance matrix, considering the test statistics of likelihood ratio, score, Wald or gradient.
- Multivariate random number generation for the multivariate t- (and gaussian) distribution.
- Graphical methods for assessing the assumption of multivariate t- (and gaussian) distribution.
- Density, distribution function and quantile function for the weights distribution arising from the multivariate t-distribution.

Please report any bugs/suggestions/improvements to Felipe Osorio, Universidad Técnica Federico Santa María. If you find these routines useful or not then please let me know. Also, acknowledgement of the use of the routines is appreciated.

Latest binaries and sources for MVT are availables from CRAN package repository .

- MVT_0.3.tar.gz - Package sources
- MVT_0.3.zip - Windows binaries (R-release)
- MVT_0.3.tgz - OS X El Capitan binaries (R-release)
- MVT_0.3.tgz - OS X Mavericks binaries (R-oldrel)
- MVT.pdf - Reference Manual

To install this package, start R and enter:

Alternatively, you can download the source as a tarball or as a zip file. Unpack the tarball or zipfile (thereby creating a directory named, MVT) and install the package source by executing (at the console prompt)

Next, you can load the package by using the command: `library(MVT)`

The package is provided under the GPL. MVT is under active development: new features are being added and old features are being improved.

Osorio, F. and Galea, M. (2017). Likelihood-based tests for the covariance matrix of multivariate t distribution: Return assessment of the Chilean pension system.

I'm Lecturer at Department of Mathematics of the Universidad Técnica Federico Santa María, Chile.

- Webpage: fosorios.mat.utfsm.cl
- Email: felipe.osorios AT usm.cl

I contribute as developer/maintainer of the following R packages:

- HEAVY - Robust estimation using heavy-tailed distributions.
- L1pack - Routines for L1 estimation.
- SpatialPack - Package for analysis of spatial data.