umap-learn - Reduce high-dimensional data with UMAP
Applies UMAP for nonlinear dimensionality reduction, visualization, supervised embeddings, and clustering preprocessing.
Tags
Updated: 2026-09-29Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Standardize input features
- Fit UMAP embeddings
- Transform new data
- Tune embedding parameters
- Use supervised labels
- Prepare clustering embeddings
- Visualize low-dimensional embeddings
Inputs
- High-dimensional data
- Feature labels
- Supervised target labels
- New input data
- UMAP parameters
Outputs
- Low-dimensional embeddings
- Transformed data embeddings
- Cluster labels
- Embedding visualizations
- Evaluation scores
Requirements
- Python environment
- umap-learn package
- scikit-learn package
- hdbscan package for clustering
- Numba support for custom metrics