umap-learn 0.5.6

Uniform Manifold Approximation and Projection

Stars: 6962, Watchers: 6962, Forks: 766, Open Issues: 455

The lmcinnes/umap repo was created 6 years ago and the last code push was 5 days ago.
The project is extremely popular with a mindblowing 6962 github stars!

How to Install umap-learn

You can install umap-learn using pip

pip install umap-learn

or add it to a project with poetry

poetry add umap-learn

Package Details

Author

None

License

BSD

Homepage

http://github.com/lmcinnes/umap

PyPi:

https://pypi.org/project/umap-learn/

GitHub Repo:

https://github.com/lmcinnes/umap

Classifiers

• Scientific/Engineering
• Software Development

Related Packages

Errors

A list of common umap-learn errors.

Code Examples

Here are some umap-learn code examples and snippets.

GitHub Issues

The umap-learn package has 455 open issues on GitHub

• [Feature Request] Allow support for "precomputed" distance matrix for umap.umap_.fuzzy_simplicial_set
• tighten test_umap_trustworthiness_fast_approx bound
• continuous UMAP latent space
• No module named 'umap'
• module 'umap' has no attribute 'UMAP'
• module 'umap' has no attribute 'version'
• Using the transform function and binary metrics for new data doesn't seem to work properly
• [Error] What if I want to apply embedding_column to category variables in Parametric UMAP?
• No conversion from array(float32, 2d, F) to array(float32, 2d, C) for 'knn_dists', defined at None
• Test failure in test_save_load with tensorflow 2.7: Unknown loss function: loss.
• Module Keras/tensorflow_probability not found when importing Parametric UMAP
• ParametricUMAP.save() compatability with Google Cloud Storage paths
• update function of UMAP does not work
• Support for wasserstein (earth mover's) distance
• LLVM ERROR: Symbol not found: __svml_sqrtf8

See more issues on GitHub