Jacknife resampling

# S4 method for FLQuant
jackknife(object, dim = "year", rm.na = TRUE)

Details

The jacknife method sets up objects ready for jacknifing, i.e. to systematically recompute a given statistic leaving out one observation at a time. From this new set of "observations" for the statistic, estimates for the bias and variance of the statstic can be calculated.

Input objects cannot have length > 1 along the iter dimension, and the resulting object will have one more iter than the number of elements in the original object.

Generic function

jacknife(object, ...)

See also

FLQuant

Examples

flq <- FLQuant(1:8) flj <- jacknife(flq) iters(flj)
#> -- iter: 1 #> 1 2 3 4 5 6 7 8 #> 1 2 3 4 5 6 7 8 #> -- iter: 2 #> 1 2 3 4 5 6 7 8 #> NA 2 3 4 5 6 7 8 #> -- iter: 3 #> 1 2 3 4 5 6 7 8 #> 1 NA 3 4 5 6 7 8 #> -- iter: 4 #> 1 2 3 4 5 6 7 8 #> 1 2 NA 4 5 6 7 8 #> -- iter: 5 #> 1 2 3 4 5 6 7 8 #> 1 2 3 NA 5 6 7 8 #> -- iter: 6 #> 1 2 3 4 5 6 7 8 #> 1 2 3 4 NA 6 7 8 #> -- iter: 7 #> 1 2 3 4 5 6 7 8 #> 1 2 3 4 5 NA 7 8 #> -- iter: 8 #> 1 2 3 4 5 6 7 8 #> 1 2 3 4 5 6 NA 8 #> -- iter: 9 #> 1 2 3 4 5 6 7 8 #> 1 2 3 4 5 6 7 NA #> #> units: NA
#> NULL
# Calculate the bias of the mean and variance estimators (mean(iter(yearMeans(flj),2:9))-c(iter(yearMeans(flj),1)))*7
#> [1] 0
(mean(iter(yearVars(flj),2:9))-c(iter(yearVars(flj),1)))*7
#> [1] 0