# 8080labs/ppscore

Predictive Power Score (PPS) in Python

Repository: https://github.com/8080labs/ppscore
Canonical: https://ross.abutalabs.com/products/ppscore
Language: Python
License: MIT
License Family: permissive
Last push: 2026-03-11T12:17:07+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 71, release rhythm 8, longevity 100
- inputs: {"age_days": 2329, "days_push": 175, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1170, forks 172 (observed 2026-08-28T04:03:51.376829+00:00)

## What it is
ppscore is a Python library implementing the Predictive Power Score (PPS), an asymmetric, data-type-agnostic metric that detects linear and non-linear relationships between dataframe columns. It serves as an alternative to correlation matrices, offering score, predictors, and matrix APIs for pandas DataFrames.

## Use cases
- find non-linear relationships between dataframe columns
- alternative to correlation matrix in pandas
- rank predictors for a target variable
- detect predictive power of features
- compute PPS matrix for feature analysis
- replace correlation with a data-type-agnostic score

## When to choose
- you need to detect non-linear or asymmetric relationships that correlation misses
- your data mixes numeric and categorical columns
- you want a quick feature-relevance ranking against a target

## When to avoid
- you need statistically rigorous dependence measures with p-values
- you work outside pandas/Python dataframes
- you need fast computation on very wide datasets, since PPS trains models per column pair

## Facets
- artifact type: library
- maturity: stable
- function: data-science, machine-learning
- domain: data-science, analytics
- platform: python
- tags: predictive-power-score, correlation-alternative, pandas, feature-analysis, feature-selection, statistics

## Member repositories
- 8080labs/ppscore (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:51.376829+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 2026-08-30T06:28:34.504965+00:00, confidence not recorded.
  - readme: https://github.com/8080labs/ppscore (fetched 2026-08-28T04:03:51.376829+00:00, sha 7d3b21ad97f7)
  - registry_pypi: https://pypi.org/pypi/ppscore/json (fetched 2026-08-29T12:34:22.468055+00:00, sha 818c75a2397d)
- Data as of 2026-08-30T08:39:29.467469+00:00.