divisio

Divisio

This gem assigns identities to a variant inside an experiment. It is similar to ab-tests framework.

A triplet - experiment, variants and identities will always produce the same variant outcome even if the test is deleted and rerun.

You can run multiple instances of the framework inside the same application for different adapters.

You can use it wherever you want however you want (you can make a global object for entire application or you can make individual little objects to be used inside methods). It is up to you do decide!

Installation

Add this line to your application's Gemfile:

gem 'divisio'

And then execute:

$ bundle

Or install it yourself as:

$ gem install divisio

Usage

No persistence adapter

This is the default adapter

It returns the selected variant, but does not persist any information.

experiment_name = 'experiment1'
variants        = [1,2,3]
identity        = 'dragos'

Divisio.new.split(experiment_name, variants, identity) # ==>> 1

Mongoid adaper

Requires mongoid v5.0.0 or greater

This adapter will persist the experiment name, identifier, and variant information in a MongoDb collection called experiments. Note: The variant returned will be cast to a string.

experiment_name = 'experiment1'
variants        = [1,2,3]
identity        = 'dragos'

Divisio.new(adapter: Divisio::MongoidAdapter).split(experiment_name, variants, identity) # ==>> '1'

You can also specify the default adapter in your initializers as so:

Divisio.default_adapter = Divisio::MongoidAdapter

then

Divisio.new.split(experiment_name, variants, identity) # ==>> '1'

Weighted variants

Variants can be weighted by passing a hash of variants mapped to their relative weights, for example:

experiment_name = 'experiment1'
variants        = { a: 1, b: 2, c: 3 }
identity        = 'dragos'

Divisio.new.split(experiment_name, variants, identity) # ==>> :c

There are three variants in this example: a, b, c

Variant a has a weight of 1, b has a weight of 2, c has a weight of 3.

Effectively, this means that the chances of getting the variant b are twice a likely as getting the variant a. The chances of getting the variant c are thrice as likely as getting the variant a.

Further examples

You can use what it returns directly, for example if you want to render a partial you could do:

partial = Divisio.new.split('amazing partial', ['partial1', 'partial2'], identity)
render(partial)

If we want to do complex logic based on some idiom, you need to do if/case statements:

# this method could be in a global helper
def enable_partial_quotes
  decision = Divisio.new.split('enable_partial_quotes', ['yes', 'no'], identity)
  decision == 'yes'
end

# .....
# then somewhere else
# .....
if enable_partial_quotes
  # do some stuff
else

Contributing

1. Fork it ( https://github.com/[my-github-username]/divisio/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