Markov chains for funand profit

About me

Jeremy Green - Consultant

Things I Enjoy : Dopamine, Serotonin

Other Interests : Drumming, Photography, and Brewing

What are Markov Chains?

According to WikiPedia

"A Markov Chain is a random process that undergoes transitions from one state to another on a state space"

State Machines

Behaving Randomly

Deterministic State Machines

Markov Chains areNon-Deterministic

"memoryless"

Applications forMarkov Chains

Moods

Stocks

Weather

Complex

Text processing

Quote Masher

An Example Deconstructed

Probability Function

How would you code this?

Starting with tests, of course!

  RSpec.describe MarkovChainer do
    before :each do
      @chainer = MarkovChainer.new
    end
    # ...
  end

  RSpec.describe MarkovChainer do
    # ...
    describe "#train" do
      it "should accept a sentence" do
        expect {
          @chainer.train("This is a test message")
        }.to_not raise_error
      end
    end
    #...
  end

  RSpec.describe MarkovChainer do
    #...
    describe "#generate with one sentece" do
      it "should return the original sentence" do
        @sentence = "This should come back un-changed."
        @chainer.train(@sentence)
        expect(@chainer.generate).to eq(@sentence)
      end
    end
    #...
  end

  RSpec.describe MarkovChainer do
    describe "#generate with two sentences" do
      it "should mix the sentences" do
        @chainer.train("Rob loves Ruby")
        @chainer.train("Jesse loves Javascript")
        results = {}
        20.times do
          new_sentence = @chainer.generate
          results[new_sentence] ||= 0
          results[new_sentence] += 1
        end
        puts results
        expect(results["Rob loves Javascript"]).to be > 0
      end
    end
  end

  # results
  {
    "Jesse loves Ruby" => 7,
    "Rob loves Ruby" => 5,
    "Rob loves Javascript" => 4,
    "Jesse loves Javascript" => 4
  }

Quote MasherAdmin Demo