# SelfExplainML/PiML-Toolbox

PiML (Python Interpretable Machine Learning) toolbox for model development & diagnostics

Repository: https://github.com/SelfExplainML/PiML-Toolbox
Canonical: https://ross.abutalabs.com/products/piml-toolbox
Homepage: https://selfexplainml.github.io/PiML-Toolbox
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: interpretable-machine-learning, low-code, ml-workflow, model-diagnostics
Last push: 2025-03-30T15:42:16+00:00

## Health v2 (maintenance only)
Score: 29/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 14, release rhythm 8, longevity 100
- inputs: {"age_days": 1587, "days_push": 521, "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 1285, forks 135 (observed 2026-08-28T04:04:14.781495+00:00)

## What it is
PiML is a Python toolbox for developing, diagnosing, and validating inherently interpretable machine learning models such as GLMs, GAMs, decision trees, EBM, GAMI-Net, XGB1/XGB2, and ReLU-DNNs, and it also wraps arbitrary black-box supervised models. It provides both a low-code graphical interface and high-code APIs covering the full model development and validation workflow, including data handling, fitting, explainability, robustness analysis, and interactive diagnostic visualizations.

## Use cases
- build interpretable machine learning models in python
- explain and diagnose ml model predictions
- validate black-box machine learning models for regulatory model risk management
- fit glm and gam models with built-in diagnostics
- low-code machine learning model development
- analyze model robustness and failure modes
- visualize feature effects and model interpretability

## When to choose
- you need inherently interpretable models (GLM, GAM, EBM, GAMI-Net, ReLU-DNN) for regulated or high-stakes decision-making
- you want one integrated workflow combining low-code UI and Python APIs for model development and validation
- you need rich diagnostic visualizations such as residual analysis, fairness, robustness, and explainability plots for tabular regression or binary classification

## When to avoid
- you are starting a new project - the authors have moved to its successor MoDeVa (modeva.ai) as of March 2025, so PiML now only receives maintenance
- you need deep learning, NLP, or LLM workflows rather than interpretable tabular supervised learning
- you only care about maximizing predictive accuracy with black-box models and have no interpretability or diagnostic requirements

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, data-science, data-visualization, monitoring
- domain: machine-learning, data-science, data-visualization, artificial-intelligence, fintech
- platform: python, cross-platform
- tags: interpretable-machine-learning, explainable-ai, model-diagnostics, model-validation, low-code, glassbox-models, model-development, xai, feature-effect-analysis, model-risk-management

## Member repositories
- SelfExplainML/PiML-Toolbox (main) score 29

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:14.781495+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 2026-08-30T04:56:38.285861+00:00, confidence not recorded.
  - readme: https://github.com/SelfExplainML/PiML-Toolbox (fetched 2026-08-28T04:04:14.781495+00:00, sha 6cbbafc70cc8)
  - homepage: https://selfexplainml.github.io/PiML-Toolbox (fetched 2026-08-29T12:12:26.039636+00:00, sha e0650c4dd418)
- Data as of 2026-08-30T08:39:29.467469+00:00.