Phitter: A library designed to streamline the process of fitting and analyzing probability distributions

This release corresponds to the version of Phitter submitted for publication in the Journal of Open Source Software (JOSS).

It includes all finalized features, documentation, and tests aligned with the JOSS submission guidelines.

Phitter v1.0.3

This release includes some bug fixes. We recommend that all users upgrade to this version.

Changes:

• The calculation of the Kolmogorov-Smirnov test statistic has been corrected. Previously, the function returned the D⁺ statistic (the upper one-tailed test statistic) now we return the D statistic.
• The correction has been applied solely to the value of the statistic; the p-value calculation remains accurate and is still based on the same theoretical foundation.

Phitter v1.0.2

Changes:

• Minor bug fixes

Phitter v1.0.1

Changes:

• Phitter uses PEP8 for all the different classes and functions

Phitter v1.0.0

This release represents a significant advancement, reflects the introduction of new features.

In addition to the library’s original capability to fit various probability distributions, it now supports:

• The simulation of processes
• The simulation of queueing systems.

This enhancement broadens the library’s utility, making it suitable for a wider range of applications in probability modeling and simulation.

Phitter v0.7.2

Release v0.7.2 - Phitter

Release Date: 28/10/2024

Overview

Phitter is a comprehensive software tool designed to analyze datasets and determine the best probability distributions that accurately represent the data. This release, version 0.7.2, introduces Phitter's core capabilities for modeling with precision and ease, providing professionals in data science, operations research, and artificial intelligence with powerful insights and interactive tools for distribution fitting.

Features

1. Extensive Probability Distribution Library

   • Phitter includes over 80 probability distributions, covering both continuous and discrete types, to ensure robust modeling for a wide range of data patterns and applications.

2. Goodness-of-Fit Tests

   • Phitter integrates 3 standardized goodness-of-fit tests that evaluate the alignment between the dataset and the proposed distributions, allowing users to select the most suitable models for their data confidently.

3. Interactive Visualizations

   • Users can leverage dynamic and interactive visualizations to explore and compare distributions, facilitating a better understanding of fit and data behavior directly within the software interface.

4. Distribution-Specific Modeling Guides

   • Each selected distribution comes with a detailed modeling guide, outlining standard methodologies and best practices, aiding users in effectively implementing the distributions in various fields.

5. Comprehensive Spreadsheets

   • Phitter provides ready-to-use spreadsheets for each distribution, documenting the underlying methodology, assumptions, and practical steps, making it easy to apply Phitter’s insights in data science, operations research, and AI projects.

System Requirements

• Operating System: Compatible with Windows, macOS, and Linux.
• Python Version: Python 3.9 or higher.
• Dependencies: Includes installation of key statistical and data visualization libraries. (Refer to the installation guide for detailed dependencies.)

Clone this version

This version can be found in branch 0.0.x