Draft:Phitter

{{AFC submission|d|adv|u=Sebastián Herrera Monterrosa|ns=118|decliner=Pythoncoder|declinets=20250617015352|ts=20250617002248}}

{{Infobox software

| name = Phitter

| logo = File:LightPhitterLogo.svg

| screenshot = File:Phitter Image.webp

| platform = Web application

| repo = {{URL|https://github.com/phitterio/phitter-kernel|GitHub Repository}}

| operating_system = Cross-platform (Windows, macOS, Linux)

| programming_language = Python

| genre = Statistical software

| language = {{hlist|English|Spanish}}

| license = MIT License

| website = {{URL|https://phitter.io|Phitter}}

}}

Phitter is an open-source Python library designed to streamline the process of fitting and

analyzing probability distributions for applications in statistics, data science, operations research, and machine learning. It provides a comprehensive catalog of over 80 continuous and

discrete distributions, multiple goodness-of-fit measures (Chi-Square, Kolmogorov-Smirnov, and

Anderson-Darling), interactive visualizations for exploratory data analysis and model validation,

and detailed modeling guides with spreadsheet implementations. By reducing the complexity

of distribution fitting, Phitter helps researchers and practitioners identify distributions that

best model their data.{{cite journal |title=Phitter: A library designed to streamline the process of fitting and analyzing probability distributions |journal=Journal of Open Source Software |date=2025 |volume=10 |number=110 |page=7625 |doi=10.21105/joss.07625 |url=https://joss.theoj.org/papers/10.21105/joss.07625 |last1=Monterrosa |first1=Sebastián José Herrera |last2=Pinilla |first2=Carlos Andrés Masmela }}{{cite web |title=Univariate Distribution Relationships |url=https://www.math.wm.edu/~leemis/chart/UDR/links.html |website=William & Mary Department of Mathematics |access-date=2025-06-16}}{{cite web |title=Phitter – A Python library for statistical distribution fitting |url=https://www.reddit.com/r/Python/comments/1hsqp3x/phitter_a_python_library_for_statistical/ |website=Reddit |date=3 January 2025 |access-date=2025-06-16}}

Features

Phitter supports fitting over 80 continuous and discrete probability distributions and includes the following features:

  • Documentations, spreadsheets and python support for continuous and discrete distributions{{cite web |title=Playground continuous and discrete distributions |url=https://phitter.io/distributions |website=Phitter |access-date=2025-06-16}}
  • Web-based interface and Python library{{cite web |title=Phitter Documentation |url=https://docs-phitter-kernel.netlify.app/ |website=Phitter Docs|access-date=2025-06-16}}
  • Goodness-of-fit tests: Chi-square, Kolmogorov–Smirnov, Anderson–Darling{{cite web |title=How to Use Goodness-of-Fit Tests to Validate Your Distribution Choice in Phitter |url=https://www.statology.org/how-to-use-goodness-of-fit-tests-to-validate-your-distribution-choice-in-phitter/ |website=Statology |date=27 February 2025 |access-date=2025-06-16}}
  • Interactive visualizations: PDF overlays, CDF plots, Q–Q plots{{cite web |title=How to Use ECDF Analysis to Validate Distribution Fits in Phitter |url=https://www.statology.org/how-to-use-ecdf-analysis-to-validate-distribution-fits-in-phitter/ |website=Statology |date=28 February 2025 |access-date=2025-06-16}}
  • Automated modeling reports with formulas and parameter estimates
  • Simulation tools for stochastic processes and queueing systems (e.g., FIFO, LIFO)
  • Parallel processing for large datasets
  • Open-source under the MIT License

See also

References