Jackknife.jl
This package provides jackknife resampling and estimation functions for Julia.
Functions
None of the functions here are exported, so you'll have to call them with the prefix
Jackknife.
or explicitly import them.
Each function takes the following two arguments:
A point estimator, given as a
Function
. The function must return a scalar when passed a vector.A real-valued vector of length > 1.
Resampling
leaveoneout(estimator, x)
Compute a vector of point estimates based on systematic subsamples of x
wherein
each index is omitted one at a time.
These are the "leave-one-out" estimates.
The resulting vector will have length length(x) - 1
.
Estimation
variance(estimator, x)
The variance of the estimator computed using the jackknife technique.
bias(estimator, x)
The bias of the estimator computed using the jackknife technique.
estimate(estimator, x)
The bias-corrected jackknife estimate of the parameter.