Class: Musikov::MarkovModel

Inherits:
Object
  • Object
show all
Defined in:
lib/musikov/markov_model.rb

Overview

This class represents a generic markov chain. It holds a hash of frequencies of subsequent states, for each state.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(value_chain = []) ⇒ MarkovModel

Initialize the hashes used to build the markov chain. Passes the initial array of values to be included in the model.

  • The “transitions” hash maps a state into another hash mapping the subsequent states to its number of subsequent occurrences

  • The “frequencies” hash maps a state into another hash mapping the subsequent states to a frequency indicating the probability of subsequent occurrences.



18
19
20
21
22
 
# File 'lib/musikov/markov_model.rb', line 18

def initialize(value_chain = [])
  @transitions = {}
  @frequencies = {}
  add_input(value_chain)
end

Instance Attribute Details

#frequencies ⇒ Object (readonly)

Returns the value of attribute frequencies.



8
9
10
 
# File 'lib/musikov/markov_model.rb', line 8

def frequencies
  @frequencies
end

Instance Method Details

#add_input(value_chain = []) ⇒ Object

Passes the argument array of state values to be included in the model.



41
42
43
44
45
46
47
48
49
50
 
# File 'lib/musikov/markov_model.rb', line 41

def add_input(value_chain = [])
  prev_value = nil
  
  value_chain.each { |value|
    add_value(prev_value, value)
    prev_value = value
  }
  
  compute_frequencies
end

#generate(initial_value, value_number) ⇒ Object

Generate a random sequence from a markov chain

  • The initial_value is a state where the random chain must start

  • The initial_value may be nil

  • The value_number indicates the number of elements on the result random sequence



28
29
30
31
32
33
34
35
36
37
38
 
# File 'lib/musikov/markov_model.rb', line 28

def generate(initial_value, value_number)
  generated_sequence = []
  selected_value = initial_value
  
  until generated_sequence.count == value_number do
    selected_value = pick_value(rand(0.1..1.0), selected_value)
    generated_sequence << selected_value
  end
  
  return generated_sequence
end