An object of the class "bw.optim" represents
a tuning parameter (usually a smoothing bandwidth)
that has been selected automatically.
The object can be used as if it were a numerical value,
but it can also be plotted to show the optimality criterion.
Adrian Baddeley Adrian.Baddeley@curtin.edu.au, Rolf Turner r.turner@auckland.ac.nz and Ege Rubak rubak@math.aau.dk.
An object of the class "bw.optim" represents the numerical
value of a smoothing bandwidth, a threshold, or a similar
tuning parameter, that has been selected by optimising
a criterion such as cross-validation.
The object is a numerical value, with some attributes that retain information about how the value was selected.
Attributes include the vector of candidate values that were examined, the corresponding values of the optimality criterion, the name of the parameter, the name of the optimality criterion, and the units in which the parameter is measured.
There are methods for print, plot,
summary, as.data.frame and as.fv
for the class "bw.optim".
The print method simply prints the numerical value of the
parameter.
The summary method prints this value, and states how
this value was selected.
The plot method produces a plot of the optimisation criterion
against the candidate value of the parameter. The as.data.frame
and as.fv methods extract this graphical information as a data
frame or function table, respectively.
Functions which produce objects of class bw.optim include
bw.CvL,
bw.CvL.adaptive,
bw.diggle,
bw.lppl,
bw.pcf,
bw.ppl,
bw.relrisk,
bw.relrisk.lpp,
bw.smoothppp
and
bw.voronoi
Ns <- if(interactive()) 32 else 3
b <- bw.ppl(redwood, srange=c(0.02, 0.07), ns=Ns)
b
summary(b)
plot(b)
Run the code above in your browser using DataLab