openTSNE: Extensible, parallel implementations of t-SNE

openTSNE is a modular Python implementation of t-Distributed Stochasitc Neighbor Embedding (t-SNE) [1], a popular dimensionality-reduction algorithm for visualizing high-dimensional data sets. openTSNE incorporates the latest improvements to the t-SNE algorithm, including the ability to add new data points to existing embeddings [2], massive speed improvements [3] [4], enabling t-SNE to scale to millions of data points and various tricks to improve global alignment of the resulting visualizations [5].

A visualization of 44,808 single cell transcriptomes obtained from the mouse retina [6] embedded using the multiscale kernel trick to better preserve the global aligment of the clusters.

User Guide

• Installation
  • Conda
  • PyPi
  • Installing from source
• How to use t-SNE
  • Simple usage
  • Advanced usage
  • Preserving global structure
  • Embedding large data sets
• How t-SNE works
  • t-SNE
  • Accelerations
  • Optimization
  • Embedding data into lower dimensions
  • References
• Parameter guide
  • Perplexity
  • Exaggeration
  • Optimization parameters
  • Barnes-Hut parameters
  • Interpolation parameters
  • References
• Benchmarks
  • Comparison to other t-SNE implementations
  • Comparison to UMAP
  • Caveats when running benchmarks
  • Reproducibility

API Reference

• API Reference
  • Initialization
  • Affinity
  • Callbacks
  • sklearn
  • OptimizationInterrupt
  • PartialTSNEEmbedding
  • TSNE
  • TSNEEmbedding

References

[1]

Van der Maaten, Laurens, and Hinton, Geoffrey. “Visualizing data using t-SNE”, Journal of Machine Learning Research (2008).

[2]

Poličar, Pavlin G., Martin Stražar, and Blaž Zupan. “Embedding to Reference t-SNE Space Addresses Batch Effects in Single-Cell Classification”, Machine Learning (2021).

[3]

Van der Maaten, Laurens. “Accelerating t-SNE using tree-based algorithms”, Journal of Machine Learning Research (2014).

[4]

Linderman, George C., et al. “Fast interpolation-based t-SNE for improved visualization of single-cell RNA-seq data”, Nature Methods (2019).

[5]

Kobak, Dmitry, and Berens, Philipp. “The art of using t-SNE for single-cell transcriptomics”, Nature Communications (2019).

[6]

Macosko, Evan Z., et al. “Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets”, Cell (2015).