tr: A interface function to use rpart() function within GAMLSS
In gamlss.add: Extra Additive Terms for Generalized Additive Models for Location Scale and Shape

Description Usage Arguments Details Value Author(s) References See Also Examples

The tr() function is a additive function to be used for GAMLSS models. It is an interface for the rpart() function of package rpart. The function tr() allows the user to use regression trees within gamlss. The great advantage of course comes from the fact GAMLSS models provide a variety of distributions and diagnostics. Note that the function gamlss.tr is not used by the user but it needed for the backfitting.

1 2	tr(formula, method = c("rpart"), control = rpart.control(...), ...) gamlss.tr(x, y, w, xeval = NULL, ...)

`formula`	A formula containing the expolanatory variables i.e. `~x1+x2+x3`.
`method`	only method "rpart" is supported at the moment
`control`	control here is equivalent to `rpart.control()` function od package `rpart`
`x`	object passing informatio to the function
`y`	the iterative y variable
`w`	the iterative weights
`xeval`	whether prediction or not is used
`...`	additional arguments

Note that, the gamlss fit maybe would not coverged. Also occasianly the gd.tol argument in gamlss has to be increased. The

Note that tr itself does no smoothing; it simply sets things up for the function gamlss() which in turn uses the function additive.fit() for backfitting which in turn uses gamlss.tr() The result is a rpart object.

Mikis Stasinopoulos mikis.stasinopoulos@gamlss.org, Bob Rigby based on work of Therneau and Atkison (2015)

Rigby, R. A. and Stasinopoulos D. M. (2005). Generalized additive models for location, scale and shape,(with discussion), Appl. Statist., 54, part 3, pp 507-554.

Rigby R.A., Stasinopoulos D. M., Heller G., and De Bastiani F., (2019) Distributions for Modeling Location, Scale and Shape: Using GAMLSS in R, Chapman and Hall/CRC.

Ripley, B. D. (1996) Pattern Recognition and Neural Networks. Cambridge.

Stasinopoulos D. M. Rigby R.A. (2007) Generalized additive models for location scale and shape (GAMLSS) in R. Journal of Statistical Software, Vol. 23, Issue 7, Dec 2007, http://www.jstatsoft.org/v23/i07.

Stasinopoulos D. M., Rigby R.A., Heller G., Voudouris V., and De Bastiani F., (2017) Flexible Regression and Smoothing: Using GAMLSS in R, Chapman and Hall/CRC.

(see also http://www.gamlss.com/).

Therneau T. M., Atkinson E. J. (2015) An Introduction to Recursive Partitioning Using the RPART Routines. Vignette in package rpart.

Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.

See Also as nn

data(rent)
#--- fitting gamlss+tree Nornal
library(rpart)
data(rent)
rg1 <- gamlss(R ~ tr(~A+Fl), data=rent, family=NO)
plot(rg1)
plot(getSmo(rg1))
text(getSmo(rg1))
## Not run: 
# fitting Gamma  errors
rg2 <- gamlss(R ~ tr(~A+Fl), data=rent, family=GA)
plot(rg2)
plot(getSmo(rg2))
text(getSmo(rg2))
#--- fitting also model in the variance
rg3 <- gamlss(R ~ tr(~A+Fl), sigma.fo=~tr(~Fl+A), data=rent, 
                family=GA, gd.tol=100, c.crit=0.1)
plot(rg3)
plot(getSmo(rg3))
text(getSmo(rg3))
plot(getSmo(rg3, what="sigma"))
text(getSmo(rg3, what="sigma"))
## End(Not run)