Why would drawing n observations without replacement from n rows make an ordinary bootstrap fail?
StatisticsLevel: FoundationBootstrap uncertainty
Why would drawing n observations without replacement from n rows make an ordinary bootstrap fail?
Hints
- What changes between resamples?
- Distinguish resampling from shuffling.
Answer
Every resample contains the same observations, so an order-invariant statistic never changes. Ordinary nonparametric bootstrap samples with replacement.
Source and licence: SciPy — resampling and Monte Carlo inference · BSD-3-Clause · answer: Quantoros editorial