Fast ways to draw from a multivariate t-distribution the scale (covariance) matrix
is sparse.

```
rmvt(n, Sigma, df = 1, delta = rep(0, nrow(Sigma)),
type = c("shifted", "Kshirsagar"), ..., sigma)
rmvt.spam(n, Sigma, df = 1, delta = rep(0, nrow(Sigma)),
type = c("shifted", "Kshirsagar"), ..., sigma)
```

## Arguments

- n
number of observations.

- Sigma
scale matrix (of class `spam`

).

- df
degrees of freedom.

- delta
vector of noncentrality parameters.

- type
type of the noncentral multivariate t distribution.

- ...
arguments passed to `rmvnorm.spam`

.

- sigma
similar to `Sigma`

. Here for portability with `mvtnorm::rmvt()`

## Details

This function is very much like `rmvt()`

from the package
mvtnorm. We refer to the help of the afore mentioned.