bsci 1.1

Constructing frequentist confidence intervals using the statistical bootstrap

Bootstrapped Confidence Intervals (bsci)

A frequentist approach to construct confidence intervals by sampling with replacement.

The statistical boostrap method was first introduced by B. Efron, "Bootstrap methods: another look at the jackknife", Annals of Statistics, 1979, link to pdf.
These lecture notes provide a nice entry-level introduction. More references are given in the code.

Citing

If you use code or ideas from this repository for your projects or research, please cite it.

@misc{Muratore_bsci,
  author = {Fabio Muratore},
  title = {bsci - Constructing frequentist confidence intervals using the statistical bootstrap},
  year = {2021},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/famura/bsci}}
}

Installation

To install the core part of the package run

pip install bsci

For (local) development install the dependencies with

pip install -e .[dev]

Getting Started

Play around with the model's parameters in the demo.py script

cd examples
python demo.py