ankane/tsne-ruby

High performance t-SNE for Ruby

t-SNE Ruby

High performance t-SNE for Ruby, powered by Multicore t-SNE

Installation

Add this line to your application’s Gemfile:

gem "tsne"

On Mac, also install OpenMP:

brew install libomp

Getting Started

Prep your data

x = [[0, 0, 0], [0, 1, 1], [1, 0, 1], [1, 1, 1]]

Run

tsne = TSNE.new
tsne.fit_transform(x)

Get Kullback-Leibler divergence

tsne.kl_divergence

Full Example

Install the matplotlib gem and download the optdigits.tes from the Optical Recognition of Handwritten Digits Data Set.

require "csv"
require "matplotlib/pyplot"
require "tsne"

data = []
target = []
CSV.foreach("optdigits.tes", converters: :numeric) do |row|
  data << row[0...-1]
  target << row[-1]
end

tsne = TSNE.new(n_jobs: 4)
embeddings = tsne.fit_transform(data)

vis_x = embeddings[true, 0]
vis_y = embeddings[true, 1]

plt = Matplotlib::Pyplot
plt.scatter(vis_x.to_a, vis_y.to_a, c: target, cmap: plt.cm.get_cmap("jet", 10), marker: ".")
plt.colorbar(ticks: 10.times.to_a)
plt.clim(-0.5, 9.5)
plt.show

Parameters

TSNE.new(
  n_components: 2,
  perplexity: 30.0,
  early_exaggeration: 12,
  learning_rate: 200,
  n_iter: 1000,
  n_iter_early_exag: 250,
  verbose: 0,
  random_state: -1,
  angle: 0.5,
  n_jobs: 1,
  cheat_metric: true
)

Data

Data can be a Ruby array

[[0, 0, 0], [0, 1, 1], [1, 0, 1], [1, 1, 1]]

Or a Numo array

Numo::DFloat.new(4, 3).rand

History

View the changelog

Contributing

Everyone is encouraged to help improve this project. Here are a few ways you can help:

• Report bugs
• Fix bugs and submit pull requests
• Write, clarify, or fix documentation
• Suggest or add new features

To get started with development:

git clone https://github.com/ankane/tsne-ruby.git
cd tsne-ruby
bundle install
bundle exec rake vendor:all
bundle exec rake test