R Graphical Manual

Browse All

Last data update: 2014.03.03

R: Plotting a simple GAMLSS model for demonstration purpose

plotSimpleGamlss

R Documentation

Plotting a simple GAMLSS model for demonstration purpose

Description

This is to plot a simple GAMLSS model where only one explanatory variable exist in order to demonstrated how the distribution of the response changes according to values of the explanatory variable.

Usage

plotSimpleGamlss(y, x, model = NULL, formula = NULL, data = NULL, 
         family = NULL, val = NULL, N = 1000, x.val = quantile(x), 
         ylim = c(min(y), max(y)), xlim = c(min(x), max(x)), ...)

Arguments

`y`	The response variable
`x`	The explanatory variable (only one is allowed here)
`model`	A fitted gamlss model
`formula`	A formula for the mean model if `model=NULL`
`data`	The data where the response and the one explanatory can be found
`family`	The gamlss family distribution
`val`	this parameter determines how the plotted distribution is shown, increase/decrease it if the distribution is not shown properly
`N`	This parameters determine how many values of y are generated for each `x.var`
`x.val`	the values of the explanatory variable where we want to see the distribution
`ylim`	the y limits in the plot
`xlim`	the x limits in the plot
`...`	extra argument to be passed to `gamlss()` function if `model=NULL`

Details

This function is for pedagogical purpose rather than fitting models to demonstrate that the distribution of the response variable can vary according to explanatory variables. In its current from it can be used with continuous and discrete responses only.

Value

A plot is shown

Author(s)

Mikis Stasinopoulos d.stasinopoulos@londonmet.ac.uk

References

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.

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.

Examples

## the abdominal data
m1 <- gamlss(y~pb(x), sigma.fo=~pb(x), data=abdom, family=LO)
plotSimpleGamlss(y,x, model=m1,   data=abdom, x.val=seq(15, 40, 5), 
                    ylim=c(0, 450), xlim=c(5, 45))
data(species)
species$ll <- log(species$lake)
m2 <- gamlss(fish~ll, data=species, trace=FALSE, family=PO )
plotSimpleGamlss(fish,ll, model=m2,   data=species, x.val=c(3,5,7, 9),
                  val=20, N=100, ylim=c(0,80))
 m3 <- gamlss(fish~ll, data=species, trace=FALSE, family=NBI, sigma.fo=~ll )
plotSimpleGamlss(fish,ll, model=m3,   data=species, x.val=c(3,5,7, 9),
                    val=20, N=100, ylim=c(0,100))
## Not run: 
##------------------------------------------------------------------------------
## the rent data 
## fitting the model first 
r1 <- gamlss(R~pb(Fl), sigma.fo=~pb(Fl),data=rent, family=GA, ylim=c(0, 3000))
## plot 1
plotSimpleGamlss(R,Fl, model=r1,   data=rent, x.val=seq(40,120, 5))
## plot 2 finer grid 
plotSimpleGamlss(R,Fl, model=r1,   data=rent, x.val=seq(40,120, 1), 
                  xlim=c(10,120))
##  the same but fitting the model within the function
## note that sigma formula has to be specified 
plotSimpleGamlss(R,Fl, formula= R~pb(Fl),   family=GA, data=rent, 
                    x.val=seq(40,120, 5), sigma.fo=~pb(Fl))
#------------------------------------------------------------------------------

## End(Not run)