A Software Tool for "Gluing" Distributions

TL;DR

This paper introduces a software tool with multiple algorithms designed to connect or 'glue' distributions from sub-intervals in Monte-Carlo simulations, aiding in large-deviation techniques.

Contribution

The paper presents a versatile software tool that implements various algorithms for effectively gluing distributions, addressing a common challenge in Monte-Carlo simulations.

Findings

Provides multiple algorithms for distribution gluing

Includes a tutorial with exemplary data

Facilitates large-deviation simulation techniques

Abstract

When performing Monte-Carlo simulations, distributions are sometimes determined only for sub-intervals of the desired total range. In such cases, a frequent problem is to connect, or glue, individual distributions to obtain the final result. Most prominent examples, where this is usually necessary, are certain large-deviation simulation techniques. However, there are multiple approaches to do this, depending on the data and individual requirements. Here, a software tool is presented, containing multiple algorithms, to aid with this task. An introduction to the available methods is presented together with a short tutorial using exemplary data.