UMAP
lmcinnesUniform Manifold Approximation and Projection, a non-linear dimension reduction method used for visualization in the way t-SNE is and for general embedding of high-dimensional data. The umap-learn package implements it with a scikit-learn interface and adds supervised and semi-supervised embedding, densMAP, AlignedUMAP for data that changes over time, and Parametric UMAP, which trains a neural network to learn the mapping.
Openness
5 high confidence- license
- BSD-3-Clause(OSI)
- source
- public(github.com/lmcinnes/umap)
- core features withheld
- no — maintained by its author
BSD-3-Clause and built from the public repository, which its author maintains. There is no company, paid tier or commercial edition behind it.
- https://raw.githubusercontent.com/lmcinnes/umap/master/LICENSE.txt recorded 2026-09-27
LICENSE.txt is the "BSD 3-Clause License", "Copyright (c) 2017, Leland McInnes".
- https://raw.githubusercontent.com/lmcinnes/umap/master/README.rst recorded 2026-09-27
README closes "If you still have questions then please open an issue and I will try to provide any help and guidance that I can."; no paid tier, hosted service or commercial license mentioned.
- https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lmcinnes/umap recorded 2026-09-27
Repository record for lmcinnes/umap, not archived and not a fork, license bsd-3-clause, last pushed 2026-09-24.
- https://ungh.cc/repos/lmcinnes/umap/files/master recorded 2026-09-27
Repository tree of 319 paths with no ee, enterprise, commercial or pro directory.
Adoption
4 high confidenceMeasured on monthly PyPI downloads of umap-learn, the package name the project publishes under.
- https://pypistats.org/api/packages/umap-learn/recent recorded 2026-09-27
last_month 5,332,757 downloads of umap-learn
Capability
2 high confidenceUMAP does one thing, embedding data in fewer dimensions, and its variants all build on that method. It is usually one step in a pipeline assembled with other tools, such as HDBSCAN for clustering the embedding.
- https://raw.githubusercontent.com/lmcinnes/umap/master/README.rst recorded 2026-09-27
README: "Uniform Manifold Approximation and Projection (UMAP) is a dimension reduction technique that can be used for visualisation similarly to t-SNE, but also for general non-linear dimension reduction."; "UMAP supports supervised and semi-supervised dimension reduction"; "Parametric UMAP provides support for training a neural network to learn a UMAP based transformation of data."
- https://umap-learn.readthedocs.io/en/latest/api.html recorded 2026-09-27
API reference documents the umap.umap_.UMAP estimator with fit, fit_transform, transform and inverse_transform.
Verified 2026-09-27