IBM::ML

A Ruby gem to invoke the IBM Machine Learning service REST API.

Currently supports:

Installation

With Gem

$ gem install ibm-ml

With Bundler

Add this line to your application's Gemfile:

gem 'ibm-ml'

And then execute:

$ bundle install

Usage

Setup

require 'ibm/ml'
require 'pp'

# input record to score 
record = {
  GENDER:        'M',
  AGEGROUP:      '45-54',
  EDUCATION:     'Doctorate',
  PROFESSION:    'Executive',
  INCOME:        200000,
  SWITCHER:      0,
  LASTPURCHASE:  3,
  ANNUAL_SPEND:  1200
}

Cloud

CLOUD_USERNAME    =  # WML service username
CLOUD_PASSWORD    =  # WML service password
CLOUD_INSTANCE_ID =  # WML instance ID
DEPLOYMENT_ID     =  # deployment ID

# Create the service object
ml_service = IBM::ML::Cloud.new(CLOUD_USERNAME, CLOUD_PASSWORD, CLOUD_INSTANCE_ID)

# Fetch an authentication token
pp ml_service.fetch_token

# Query models
pp ml_service.models
pp ml_service.model_by_name('ML Model')

# Query deployments
pp ml_service.deployments
pp ml_service.deployment(DEPLOYMENT_ID)                 #    identify by deployment ID
pp ml_service.deployment_by_name('Deployed ML Model')   # OR identify by deployment name

# Get a score for the given deployment and record
score = ml_service.score(DEPLOYMENT_ID, record)                 #    identify by deployment ID
score = ml_service.score_by_name('Deployed ML Model', record)   # OR identify by deployment name

Local

LOCAL_HOST      =  # DSX Local hostname / IP address
LOCAL_USERNAME  =  # DSX Local username
LOCAL_PASSWORD  =  # DSX Local password
DEPLOYMENT_ID   =  # deployment ID

# Create the service object
ml_service = IBM::ML::Local.new(LOCAL_HOST, LOCAL_USERNAME, LOCAL_PASSWORD)

# Fetch an authentication token
pp ml_service.fetch_token

# Get a score for the given deployment and record
score = ml_service.score(DEPLOYMENT_ID, record)

Print Score

pp score  # print full score hash

# extract prediction and probability from score
prediction = ml_service.query_score(score, 'prediction')
probability = ml_service.query_score(score, 'probability')[prediction]

puts
puts "Prediction = #{prediction == 1}"                  # print binary 1/0 prediction as true/false
puts "Probability = #{(probability * 100).round(1)}%"   # print probability value as percentage

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/IBM-DSE/ibm-ml-ruby.