Compute one draw of horseshoe prior parameters (local precision, global
precision, and local and global parameter expansion terms).
Usage
sampleHS(omega, params, sigma_e = 1)
Arguments
- omega
n x p matrix of errors
- params
list of parameters to update
- sigma_e
the observation error standard deviation; for (optional) scaling purposes
Value
List of relevant components in params: sigma_wt, the n x p matrix of standard deviations,
and the local and global precisions and parameter-expansion terms.
Note
To avoid scaling by the observation standard deviation sigma_e,
simply use sigma_e = 1 in the functional call.