# astroautomata/PySR

High-Performance Symbolic Regression in Python and Julia

Repository: https://github.com/astroautomata/PySR
Canonical: https://ross.abutalabs.com/products/pysr
Homepage: https://ai.damtp.cam.ac.uk/pysr
Language: Python
License: Apache-2.0
License Family: permissive
Topics: symbolic-regression, machine-learning, python, julia, genetic-algorithm, automl, interpretable-ml, data-science, explainable-ai, scikit-learn, symbolic, algorithm, distributed-systems, evolutionary-algorithms, equation-discovery
Last push: 2026-08-26T18:12:18+00:00

## Health v2 (maintenance only)
Score: 99/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 100
- inputs: {"age_days": 2179, "days_push": 7, "days_rel": 7, "gap_med": 5, "n_releases_24m": 20}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3739, forks 349 (observed 2026-08-28T04:08:15.905321+00:00)

## What it is
PySR is a high-performance symbolic regression library that searches for symbolic mathematical expressions optimizing a given objective, implemented in Julia with a Python interface. It integrates with scikit-learn and supports distributed, evolutionary search for interpretable equation discovery.

## Use cases
- discover symbolic equations from data
- fit interpretable mathematical models to datasets
- perform symbolic regression in Python
- find closed-form expressions approximating noisy data
- use evolutionary algorithms for equation discovery
- explainable machine learning model discovery

## When to choose
- you need interpretable, human-readable equations instead of black-box models
- you want high-performance symbolic regression with distributed search
- you work in Python with scikit-learn-style APIs
- you need equation discovery for scientific data

## When to avoid
- you need maximum predictive accuracy regardless of interpretability
- your data is best modeled by deep neural networks
- you need a pure-Python dependency without Julia runtime

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, data-science, math
- domain: machine-learning, data-science
- platform: python, jvm-scripting, windows
- tags: symbolic-regression, julia, genetic-algorithms, interpretable-ml, explainable-ai, equation-discovery, scikit-learn, automl, evolutionary-algorithms, algorithms, linux, macos, docker

## Member repositories
- astroautomata/PySR (main) score 99

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:15.905321+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:29:08.401823+00:00, confidence not recorded.
  - readme: https://github.com/astroautomata/PySR (fetched 2026-08-28T04:08:15.905321+00:00, sha af18329e0201)
  - homepage: https://ai.damtp.cam.ac.uk/pysr (fetched 2026-08-29T09:23:28.729011+00:00, sha afb009270d3a)
- Data as of 2026-08-30T08:39:29.467469+00:00.