# PAIR-code/facets

Visualizations for machine learning datasets

Repository: https://github.com/PAIR-code/facets
Canonical: https://ross.abutalabs.com/products/facets
Homepage: https://pair-code.github.io/facets/
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: machine-learning, data-visualization
Archived: true
Last push: 2023-05-24T15:56:22+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3344, "days_push": 1197, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7336, forks 883 (observed 2026-08-28T04:09:58.306641+00:00)

## What it is
Facets is a pair of web-component visualizations (Overview and Dive) for understanding and analyzing machine learning datasets, embeddable in Jupyter notebooks or web pages. Overview provides feature-by-feature statistics and dataset comparison, while Dive enables interactive exploration of individual data points.

## Use cases
- visualize feature distributions in an ML dataset
- compare training and test set statistics to detect skew
- find missing values and unexpected feature values in a dataset
- interactively explore tens of thousands of data points in a notebook
- detect training/serving skew before deploying a model
- explore outliers in multidimensional data

## When to choose
- you need quick feature-level statistics and distribution comparisons for ML datasets
- you work in Jupyter notebooks and want embeddable dataset visualizations
- you want to spot dataset issues like missing values or train/test skew

## When to avoid
- you need general-purpose business intelligence dashboards
- you need actively developed tooling with frequent updates
- you need visualization of model performance rather than datasets

## Facets
- artifact type: library
- maturity: maintenance
- function: data-visualization, machine-learning, analytics
- domain: machine-learning, data-visualization, data-science
- platform: python, browser, cross-platform
- tags: jupyter-notebook, web-components, dataset-analysis, exploratory-data-analysis, pair, web

## Member repositories
- PAIR-code/facets (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:58.306641+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 2026-08-29T17:38:28.663768+00:00, confidence not recorded.
  - readme: https://github.com/PAIR-code/facets (fetched 2026-08-28T04:09:58.306641+00:00, sha f2e6ff87f4c0)
  - homepage: https://pair-code.github.io/facets/ (fetched 2026-08-29T08:33:57.816094+00:00, sha 6d9585433742)
- Data as of 2026-08-30T08:39:29.467469+00:00.