README.md

☗ Mikon

Mikon is a flexible data structure for Ruby language, inspired by data.frame of R and Pandas of Python. Its goal is to make it easy to manipulate the real data, apply statistical function to it and visualize the result in Ruby language.

It is compatible with Nyaplot::DataFrame and Statsample::Vector, and most methods the both gem have can be applied to Mikon's data structure.

Main Features:

• Fast data manipulation with NMatrix
• Advanced plotting with Nyaplot

Dependencies

• CRuby >= 2.0.0-p451
• NMatrix >= v0.1.0.rc5
• Formatador >= 0.2.5

Optional Dependencies

• Nyaplot: for plotting
• Statsample: for statistical function
• IRuby: for the interactive manipulation of data

Installation

$ gem install mikon

If you fail to install nmatrix, try:

$ sudo apt-get install libatlas-base-dev
$ sudo apt-get --purge remove liblapack-dev liblapack3 liblapack3gf
$ gem install nmatrix -- --with-opt-include=/usr/include/atlas

More detailed instructions are available for Mac and Linux.

If you fail to install iruby, try this.

Examples

Notebooks created with IRuby:

• Basic Data Manipulation
• Statistical functions
• Plotting

Usage

Initializing DataFrame

require 'mikon'
df2 = Mikon::DataFrame.new([{a: 1, b: 2}, {a: 2, b: 3}, {a: 3, b: 4}])

Mikon::DataFrame.new({a: [1,2,3,4], b: [2,3,4,5]}, index: [:a, :b, :c, :d])

df = Mikon::DataFrame.from_csv("~/data.csv")

Basic data manipulating

df[:value]

df[10..20]

df.head(2)

df.tail(2)

Row-based data manipulating

df.select{value > 100}

df2.map{b+1}.name(:c)

foo = []
df.each{foo.push(2*a)}
p foo #-> [2,4,6]

df.insert_column(:new_value){value * 2}

df.any?{value >= 100} #-> true
df.all?{valu > 1} #-> false

Column-based data manipulating

In most cases column-based manipulating is faster than Row-based.

df2[:b] - df2[:a]

df.insert_column(:new_value, df[:value]*2)

Plotting

df[:value].plot

Plotting with Nyaplot

require 'nyaplot'
plot = Nyaplot::Plot.new
plot.add_with_df(df, :histogram, :value)
plot

Statistical with Statsample

Mikon::Series is compatible with Statsample::Vector, so most methods of Statsample can be applied to Mikon::Series.

require 'statsample'

Statsample::Analysis.store(Statsample::Test::T) do
  t_2 = Statsample::Test.t_two_samples_independent(df1[:value], df1[:new_value])
  summary t_2
end

Statsample::Analysis.run_batch

License

MIT License

Acknowledgement

Ruby Association Grant 2014 has been earmarked for the development of Mikon.

Contributing

1. Fork it ( http://github.com/domitry/mikon/fork )
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Run tests by running rspec on /path_to_gem/mikon/
5. Push to the branch (git push origin my-new-feature)
6. Create new Pull Request