In Julia, we have currently implemented Rao and Wu's (n-1) bootstrap function called bootweights() in src/bootstrap.jl (confusingly, this is called subbootstrap in R survey). The method corresponding to bootstrap in R is Canty and Davison's bootstrap.
Most of the processing steps are same between Rao-Wu and Canty Davison, except subtle differences when complete population information and finite parameter correction (fpc) is given, and the way scaling is done in line 71 vs 101. But since we dont have fpc corrections yet in Julia, so for our case should be minor changes from the current src/bootweights() function
See line 23, R survey/bootstrap.R for analogous code in R.
In Julia, we have currently implemented Rao and Wu's (n-1) bootstrap function called
bootweights()
insrc/bootstrap.jl
(confusingly, this is calledsubbootstrap
in R survey). The method corresponding tobootstrap
in R is Canty and Davison's bootstrap.Most of the processing steps are same between Rao-Wu and Canty Davison, except subtle differences when complete population information and finite parameter correction (fpc) is given, and the way scaling is done in line 71 vs 101. But since we dont have fpc corrections yet in Julia, so for our case should be minor changes from the current
src/bootweights()
function