R/AllGeneric.R

In vcrpart: Tree-Based Varying Coefficient Regression for Generalized Linear and Ordinal Mixed Models

## --------------------------------------------------------- #
## Author:          Reto Buergin
## E-Mail:          rbuergin@gmx.ch
## Date:            2016-01-10
##
## Description:
## New generics for exported methods
##
## Overview:
## cvloss:             cross-validates the loss of a model
##                     regarding a tuning parameter. See
##                     'tvcm-cv.R'.
## extract:            extracts features of a fitted model.
##                     See 'tvcm-methods.R'.
## neglogLik2:         extracts the -2*log-likelihood loss of
##                     a fitted model. See also 'olmm-methods.R'
##                     and 'tvcm-methods.R'
## neglogLik2.default: neglogLik2 function which should
##                     work for most model classes
## oobloss:            estimates out-of-bag loss.
## otsplot:            ordinal time series plot.
## prunepath:          extracts the path of pruning a fitted
##                     tree.
## ranefCov:           extracts covariance matrix of random
##                     effect variance parameters.
## splitpath:          extracts the splitting path of the
##                     growing process of a tree.
##
## Last modifications:
## 2014-09-07: added 'prunepath' generic
## 2014-07-17: Update the descriptions
## --------------------------------------------------------- #

cvloss <- function(object, ...) UseMethod("cvloss")

extract <- function(object, ...) UseMethod("extract")

fixef.glm <- function(object, ...) coef(object)

neglogLik2 <- function(object, ...) UseMethod("neglogLik2")

neglogLik2.default <- function(object, ...)
    return(-as.numeric(2 * logLik(object)))

oobloss <- function(object, ...) UseMethod("oobloss")

otsplot <- function(x, ...) UseMethod("otsplot")

prunepath <- function(tree, ...) UseMethod("prunepath")

ranefCov <- function(object, ...) UseMethod("ranefCov")

splitpath <- function(tree, ...) UseMethod("splitpath")