Class: MarkovChain

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

Defined Under Namespace

Modules: Prediction

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(data = {}) ⇒ MarkovChain

creates an empty MarkovChain.

data is a map which becomes owned by this MarkovChain.



17
18
19
20
 
# File 'lib/markov_chain.rb', line 17

def initialize(data = {})
  @data = data
  @last_state = nil
end

Class Method Details

.from(data) ⇒ Object

data is #data() of the other MarkovChain. It becomes owned by the returned MarkovChain.



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

def self.from(data)
  new(data)
end

Instance Method Details

#append!(states) ⇒ Object

appends states to the end of this MarkovChain.

states is an Array of arbitrary objects.

It returns this (modified) MarkovChain.



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

def append!(states)
  for next_state in states
    state_occurences_map = (@data[@last_state] or Hash.new)
    state_occurences_map[next_state] ||= 0
    state_occurences_map[next_state] += 1
    @data[@last_state] = state_occurences_map
    @last_state = next_state
  end
  return self
end

#data ⇒ Object

data passed to MarkovChain.new() or MarkovChain.from().



58
59
60
 
# File 'lib/markov_chain.rb', line 58

def data
  @data
end

#predict ⇒ Object

returns Enumerable of predicted states. The states are predicted by states passed to #append!().

The result may contain nils. Each nil means that MarkovChain could not predict a state after the one before nil. Example:

markov_chain.predict().take(4)  #=>  ["a", "c", "b", nil]

That means markov_chain could not predict a state after “b”.



51
52
53
 
# File 'lib/markov_chain.rb', line 51

def predict()
  self.extend(Prediction)
end