resamp
. For each parameter, a histogram of replicates
with a smooth density estimate is plotted. A solid line indicates the observed
parameter value for the original data, and a dotted line indicates the mean of
the replicates. The difference between these two lines is the estimated bias.
plot.resamp(x, nrow = NULL, grid.layout = T, rugplot = F, nclass.func = nclass.FD, bandwidth.func = bandwidth.nrd, ...)
resamp
such as is produced by
bootstrap
or
jackknife
.
nrow
is NULL and
x
has a
dim.obs
component, it is used to determine
nrow
. Otherwise, two rows are
used unless there is only a single parameter.
nclass.FD
,
which is defined within
plot.resamp
, is used. This function uses the
Freedman and Diaconis rule as described in Venables and Ripley.
density
when
constructing the density estimate. By default, the function
bandwidth.nrd
is used. This function uses the
normal reference density bandwidth estimate as described by Venables and
Ripley.
hist
and
barplot
.
Venables, W. N. and Ripley, B. D. (1994). Modern Applied Statistics with S-PLUS. New York: Springer-Verlag.
temp <- bootstrap(stack.loss,var) plot(temp)