Fitting Semi-parametric Symmetric Regression Models
Description
This package allows to fit a regression model suitable for data analysis in which the conditional distribution of the response is symmetric continuous and where its location and dispersion parameters are explicitly modelled through semi-parametric functions, whose non-parametric components are approximated by natural cubic splines or P-splines. Supported distributions for the model error include normal, Student-t, power exponential, symmetric hyperbolic, contaminated normal, slash, sinh-normal and sinh-t distributions.