resamp
objects. This
calculates percentiles for the
parameter replicates, and correlations of parameter replicates
if applicable. The print method displays the same results as
,
plus these quantities.
summary.resamp(object, probs = c(0.025, 0.05, 0.95, 0.975), narrow = F)
resamp
.
narrow=FALSE
(the default), then intervals are wider and coverage probabilities
are usually more accurate than with
narrow=TRUE
.
object
(
call
,
B
,
observed
,
estimate
), plus percentiles
limits.emp
and, if
observed
has more than one element,
the correlation matrix
correlation
for the replicates.
The print method prints the results in a nicely formatted manner.
summary(jackknife(stack.loss,var))