Git Product home page Git Product logo

pca's Introduction

Principal Component Analysis (PCA)

Principal component analysis in Ruby. Uses GSL for calculations.

PCA can be used to map data to a lower dimensional space while minimizing information loss. It's useful for data visualization, where you're limited to 2-D and 3-D plots.

For example, here's a plot of the 4-D iris flower dataset mapped to 2-D via PCA:

iris

PCA is also used to compress the features of a dataset before feeding it into a machine learning algorithm, potentially speeding up training time with a minimal loss of data detail.

Install

GSL must be installed first. On OS X it can be installed via homebrew: brew install gsl

gem install pca

Example Usage

require 'pca'

pca = PCA.new components: 1

data_2d = [ 
  [2.5, 2.4], [0.5, 0.7], [2.2, 2.9], [1.9, 2.2], [3.1, 3.0],
  [2.3, 2.7], [2.0, 1.6], [1.0, 1.1], [1.5, 1.6], [1.1, 0.9]
]

data_1d = pca.fit_transform data_2d

# Transforms 2d data into 1d:
# data_1d ~= [
#   [-0.8], [1.8], [-1.0], [-0.3], [-1.7],
#   [-0.9], [0.1], [1.1], [0.4], [1.2]
# ]

more_data_1d = pca.transform [ [3.1, 2.9] ]

# Transforms new data into previously fitted 1d space:
# more_data_1d ~= [ [-1.6] ]

reconstructed_2d = pca.inverse_transform data_1d

# Reconstructs original data (approximate, b/c data compression):
# reconstructed_2d ~= [
#   [2.4, 2.5], [0.6, 0.6], [2.5, 2.6], [2.0, 2.1], [2.9, 3.1]
#   [2.4, 2.6], [1.7, 1.8], [1.0, 1.1], [1.5, 1.6], [1.0, 1.0]
# ]

evr = pca.explained_variance_ratio

# Proportion of data variance explained by each component
# Here, the first component explains 99.85% of the data variance:
# evr ~= [0.99854]

See examples for more. Also, peruse the source code (~ 100 loc.)

Options

The following options can be passed in to PCA.new:

option default description
:components nil number of components to extract. If nil, will just rotate data onto first principal component
:scale_data false scales features before running PCA by dividing each feature by its standard deviation.

Working with Returned GSL::Matrix

PCA#transform, #fit_transform, #inverse_transform and #components return instances of GSL::Matrix.

Some useful methods to work with these are the #each_row and #each_col iterators, and the #row(i) and #col(i) accessors.

Or if you'd prefer to work with a standard Ruby Array, you can just call #to_a and get an array of row arrays.

See GSL::Matrix RDoc for more.

Plotting Results With GNUPlot

Requires GNUPlot and gnuplot gem.

require 'pca'
require 'gnuplot'

pca = PCA.new components: 2
data_2d = pca.fit_transform data

Gnuplot.open do |gp|
  Gnuplot::Plot.new(gp) do |plot|
    plot.title "Transformed Data"
    plot.terminal "png"
    plot.output "out.png"

    # Use #col accessor to get separate x and y arrays
    # #col returns a GSL::Vector, so be sure to call #to_a before passing to DataSet
    xy = [data_2d.col(0).to_a, data_2d.col(1).to_a]

    plot.data << Gnuplot::DataSet.new(xy) do |ds|
      ds.title = "Points"
    end
  end
end

Sources and Inspirations

pca's People

Contributors

gbuesing avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar

pca's Issues

Saving eigenvectors and eigenvalues?

Hi,

Is it possible to save the eigenvectors and eigenvalues for a PCA object such that they could be re-loaded and used when new vectors are added to a dataset?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.