largeVis

This is an implementation of the largeVis algorithm described in (https://arxiv.org/abs/1602.00370). It also incorporates:

• A very fast algorithm for estimating k-nearest neighbors, implemented in C++ with Rcpp and RcppParallel, using RcppAnnoy for the ANNOY phase of the algorithm.
• Efficient implementations of the clustering algorithms:
  • HDBSCAN
  • OPTICS
  • DBSCAN
• Functions for visualizing manifolds like this.

News Highlights

• Version 0.3.0 is a major update. OpenMP has been replaced by RcppParallel. My homegrown implementation of the ANNOY phase of the ANN algorithm is replaced by RcppAnnoy. This was done because of the varying availability of OpenMP on many systems (e.g., OS X). Performance appears to be improved considerably, and more distance metrics are available, unfortunately at the expense of breaking changes to several functions. See NEWS for full details.
• Version 0.1.10 re-adds clustering, and also adds momentum training to largeVis, as well as a host of other features and improvements.
• Version 0.1.9.1 has been accepted by CRAN. Much grattitude to Uwe Ligges and Kurt Hornik for their assistance, advice, and patience.

Some Examples

<figure> <img src="./README_files/figure-markdown_github/drawmnist-1.png" alt="MNIST" /> <figcaption aria-hidden="true">MNIST</figcaption> </figure> <figure> <img src="./README_files/figure-markdown_github/drawwikiwords-1.png" alt="Wiki Words" /> <figcaption aria-hidden="true">Wiki Words</figcaption> </figure>

Clustering With HDBSCAN

<!-- -->

Visualize Embeddings

<figure> <img src="./README_files/figure-markdown_github/faceImages-1.png" alt="Visualize Embeddings" /> <figcaption aria-hidden="true">Visualize Embeddings</figcaption> </figure>