Digiproc

Use Digiproc to perform a number of digital signal processing calculations. Easily represent digital and analog signals, find their Discrete Fourier Transform in O(n*lgn) time, and create various systems from filtering, modulation, encoding, etc. Digiproc also enables you to easily solve a wide range of probability and linear algebra-based problems.

Easily visualize data in a graphing API built on top of gruff.

See the examples folder as well as the specs.

Installation

Add this line to your application's Gemfile:

gem 'digiproc'

And then execute:

$ bundle

Or install it yourself as:

$ gem install digiproc

If you are having trouble downloading the rmagick gem on OSX, try the following:

    brew uninstall imagemagick
    brew install imagemagick@6
    export PATH="/usr/local/opt/imagemagick@6/bin:$PATH"
    brew link --force imagemagick@6
    gem install rmagick

Usage

Graphing capabillity built on top of gruff gem, which uses the rMagick gem and ImageMagick.

Look at specs and the examples folder for examples on how to use these classes in calculations

Limitations to be aware of:

  • All signals must be causual -> negative values of n are not supported
  • The FFT, as of now, uses a Radix 2 algorithm. So, the overall size must be a power of 2 - any dataset is zero-filled automatically to meet this necessity. This gives you less control over the exact size if needed and may cause slower runtimes (however the radix 2 algorithm is ϴ(nlgn)). This zero-fill will increase the resolution of the FFT output, however.
  • All filter data and windows must have an even number of datapoints. If not, this will be done automatically. This is to facilitate the ability to create any type of filter using the windowing method. (Odd numbers of values can preclude certain types of filters)
  • All filters are FIR and are implemented via the windowing method. Although you can get pretty good results depending on your application, this is an older method compared to the optimal Parks-McClellan algorithm.
  • IIR Filter design is not included
  • The quantization process in Digiproc::AnalogSignal maps an analog signal to -1 to 1 before quantizing the result. A floating point number is accurate up to 7 decimal places, so this process will cause unwanted ACTUAL quantization errors (not simulated quantization errors) for an amplitude range greater than $2 x 10^16$ (ie 20 Quadrillion)

This is still a work in progress. There are many usuable functions, and many that still require tuning. Functions tested in the spec file are very reliable.

Run Examples

Use rake examples:run[file] to run an example file to see how it works, ie: rake examples:run[binomial_distribution/dice]

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/micahshute/digiproc.

There is a long list of intended actions in the TODO.md file. Tests (rspec) covers many core capabilities, but more need to be written to cover the entire gem.

This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct. That being said, negative feedback is always welcome.

License

The gem is available as open source under the terms of the MIT License.

Code of Conduct

Everyone interacting in the Digiproc project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.