Rubadana

Rubadana is an elementary ruby data-analysis package. It works with plain old ruby objects, not sql or databases or anything fancy like that.

The aim is to create a summary overview from a list of objects by basically running a group-by/map/reduce operation on your list. The input, grouping, mapping, reducing, and display of the result are all independently variable.

Installation

Add this line to your application's Gemfile:

gem 'rubadana'

And then execute:

$ bundle

Or install it yourself as:

$ gem install rubadana

Usage

Here's a trivial example, returning the number of requests per user:

  program  = Rubadana::Registry.build groupings: [:user], mappers: [:self], reducers: [:count]
  analysis = program.run Request.all

In this example, :user, :self, and :count are plugins you will have provided to rubadana for extracting and manipulating your data.

analyser.run returns a list of Rubadana::Analysis instances, with, for this example, the following attributes:

|key | a Hash instance with keys :user | |list | the subset of Request.all with the corresponding value for :user | |mapped | in this case, the same as list (assuming the :identity mapper returns the thing itself) | |reduced| a list of one integers, equal to the size of the list |

In ordinary ruby, you would write Request.all.group_by(:user).map {|user, requests| [user, requests.count] } to get the same information.

Here's a richer example which returns the sum of debits, credits, and account balances from a set of accounting transactions:

  program = Rubadana::Analyser.new group: [:month, :account_number], map: [:debits, :credits, :balance], reduce: [:sum, :sum, :sum]
  analysis = program.run AccountingTransaction.all

In this example, :month, :account_number, :debits and so on, are plugins you will have provided to rubadana for extracting and manipulating your data.

analyser.run returns a list of Rubadana::Analysis instances, with the following attributes:

|key | a Hash instance with keys :month and :account_number | |list | the subset of AccountingTransaction.all having the corresponding values for :month and :account_number | |mapped | the output of the map operations on list. This is a list of n-tuples, where n is the number of operations specified by the map parameter | |reduced| the output of the reduce operations on mapped. This is a list of n values, one for each operation specified by the reduce parameter. |

In this example, reduced gives us the sum of all debits, the sum of all credits, and the sum of all balances, per account-number

See spec for some examples.

Steps

1. Create a Registry for your mappers and your reducers

my_registry = Rubadana::Registry.new

1. Create and register some mappers

  class SaleYear
    def name        ; :yearly         ; end
    def run  thing  ; thing.date.year ; end
    def label value ; value           ; end
  end

  my_registry.register_mapper SaleYear.new

1. Create and register some reducers:

  class Sum
    def name          ; :sum             ; end
    def reduce things ; things.reduce :+ ; end
  end

  my_registry.register_reducer Sum.new

1. Build an analysis program and run it:

  # this is a program to analyse invoices by year and product, giving the
  # number of sales, the sum of sales and the average sale in each case
  my_program = register.build group: %i{ yearly }, map: %i{ self sale_amount sale_amount }, reduce: %i{ count sum average }

  data = my_program.run(invoices)

#run returns an array of Rubadana::Analysis as described above.

Contributing

1. Fork it ( https://github.com/conanite/rubadana/fork )
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request