Fast Tail Index Estimation for Power Law Distributions in R

TL;DR

This paper introduces an R package for fast, accurate tail index estimation of power law distributions, especially suited for large datasets, emphasizing ease of use and speed.

Contribution

The paper presents a new R package that significantly improves the speed of tail index estimation for power law distributions while maintaining accuracy.

Findings

The package offers faster estimation compared to existing methods.

It maintains high accuracy in tail index estimation.

Demonstrates effectiveness on large datasets.

Abstract

Power law distributions, in particular Pareto distributions, describe data across diverse areas of study. We have developed a package in R to estimate the tail index for such datasets focusing on speed (in particular with large datasets), keeping in mind ease of use, as well as accuracy. In this document, we provide a user guide to our package along with the results obtained highlighting the speed advantages of our package.