Module bootstrap 

The bootstrap (ISL ch 5.2) — resample-with-replacement to estimate the sampling distribution (standard error / spread) of an arbitrary statistic.

Structs

BootstrapCi

An earned confidence interval — derived by resampling the data, not assumed from a Gaussian.

BootstrapResult

Bootstrap estimate of a scalar statistic’s sampling distribution.

Enums

CiMethod

Which bootstrap confidence-interval to compute.

Functions

bootstrap_ci

Bootstrap confidence interval for statistic at confidence 1 − alpha (alpha in (0,1), e.g. 0.05 → 95%). None if data is empty, b < 2, or alpha is out of range.

bootstrap_estimate

Estimate the standard error (and bias) of statistic applied to data, over b bootstrap resamples. statistic maps a sample slice to a scalar. None if data is empty or b < 2.

bootstrap_indices

One bootstrap resample of n row indices, drawn with replacement.