rubykov 0.2.0 → 0.3.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/rubykov.rb +2 -0
- data/lib/rubykov/hidden_markov_model.rb +58 -0
- data/lib/rubykov/word_generator.rb +44 -0
- metadata +5 -3
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: be1de15771cfa1694a274b2ad2659c565c0898f1
|
4
|
+
data.tar.gz: 1cc4d87fbacfeeaeebf0ad8c32e15cff7887e480
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 0fcc4cd9b24f2185c2ee429fbf37de51c0ef8ec0efd5c117afb42632a0d34662131e48c056b01496d682f663ea636cee1b5774b1ea89ca1d06f854de5e19236b
|
7
|
+
data.tar.gz: bfa02c7f821816395aaf2266fdb85cd94ce798453121fda7d73ebd3fb3912ec9b44c48e8b6684920a3ac7cff20d55ef0ffd983644ae99b24da980d987434aed3
|
data/lib/rubykov.rb
CHANGED
@@ -0,0 +1,58 @@
|
|
1
|
+
module Rubykov
|
2
|
+
class HiddenMarkovModel
|
3
|
+
def initialize(states, observations, start_probability, transition_probabilities, emission_probabilities)
|
4
|
+
@states = states
|
5
|
+
@observations = observations
|
6
|
+
@start_probability = start_probability
|
7
|
+
@transition_probabilities = transition_probabilities
|
8
|
+
@observation_probabilities = emission_probabilities
|
9
|
+
end
|
10
|
+
|
11
|
+
def most_likely_state_sequence(observation_sequence)
|
12
|
+
max_probabilities(observation_sequence).map(&:first)
|
13
|
+
end
|
14
|
+
|
15
|
+
def most_likely_state_sequence_probability(observation_sequence)
|
16
|
+
max_probabilities(observation_sequence).last.last
|
17
|
+
end
|
18
|
+
|
19
|
+
private
|
20
|
+
|
21
|
+
def max_probabilities(observation_sequence)
|
22
|
+
compute_viterbi_probabilities(observation_sequence.map(&:to_sym)).map do |probability|
|
23
|
+
probability.max_by { |k,v| v }
|
24
|
+
end
|
25
|
+
end
|
26
|
+
|
27
|
+
def compute_viterbi_probabilities(observations)
|
28
|
+
state_max_probabilities = []
|
29
|
+
|
30
|
+
observations.each_with_index do |observation, index|
|
31
|
+
state_max_probabilities[index] = {}
|
32
|
+
@states.each do |state|
|
33
|
+
state_max_probabilities[index][state] = 0
|
34
|
+
state_max_probabilities.first[state] = @start_probability[state] * @observation_probabilities[state][observations.first]
|
35
|
+
end
|
36
|
+
end
|
37
|
+
|
38
|
+
observations[1..-1].each_with_index do |observation, index|
|
39
|
+
@states.each do |state|
|
40
|
+
max_probability = state_max_probabilities[index].map { |previous_state, previous_probability| @transition_probabilities[previous_state][state] * previous_probability }.max
|
41
|
+
state_max_probabilities[index+1][state] = @observation_probabilities[state][observation] * max_probability
|
42
|
+
end
|
43
|
+
end
|
44
|
+
|
45
|
+
state_max_probabilities
|
46
|
+
end
|
47
|
+
|
48
|
+
def sample(probability_distribution)
|
49
|
+
random_value = rand
|
50
|
+
|
51
|
+
probability_ranges = probability_distribution.inject([0]) do |memo, p|
|
52
|
+
memo << p.last + memo.last
|
53
|
+
end
|
54
|
+
|
55
|
+
probability_distribution[probability_ranges.rindex { |x| x <= random_value }].first
|
56
|
+
end
|
57
|
+
end
|
58
|
+
end
|
@@ -0,0 +1,44 @@
|
|
1
|
+
module Rubykov
|
2
|
+
class WordGenerator < MarkovModel
|
3
|
+
def initialize(order, training_data)
|
4
|
+
super(order, training_data)
|
5
|
+
end
|
6
|
+
|
7
|
+
def train(training_data)
|
8
|
+
add_data_to_model(training_data)
|
9
|
+
end
|
10
|
+
|
11
|
+
def word
|
12
|
+
chain.to_a.join('').capitalize
|
13
|
+
end
|
14
|
+
|
15
|
+
private
|
16
|
+
|
17
|
+
def massage_training_data(training_data)
|
18
|
+
training_data.map { |datum| datum.downcase.split('') + ['\n'] }
|
19
|
+
end
|
20
|
+
|
21
|
+
def add_data_to_model(training_data)
|
22
|
+
massage_training_data(training_data).each do |datum|
|
23
|
+
super(datum)
|
24
|
+
end
|
25
|
+
end
|
26
|
+
|
27
|
+
def chain_enumerator(seed_state = states.sample)
|
28
|
+
Enumerator.new do |output|
|
29
|
+
current_state = seed_state
|
30
|
+
current_state.each do |word|
|
31
|
+
output << word
|
32
|
+
end
|
33
|
+
|
34
|
+
loop do
|
35
|
+
break if @representation[current_state].nil?
|
36
|
+
next_letter = @representation[current_state].sample
|
37
|
+
break if next_letter == '\\n'
|
38
|
+
output << next_letter
|
39
|
+
current_state = current_state.last(@order-1) + [next_letter]
|
40
|
+
end
|
41
|
+
end
|
42
|
+
end
|
43
|
+
end
|
44
|
+
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: rubykov
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.
|
4
|
+
version: 0.3.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Evan Hemsley
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2017-07-24 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: rspec
|
@@ -38,8 +38,10 @@ extra_rdoc_files: []
|
|
38
38
|
files:
|
39
39
|
- README.md
|
40
40
|
- lib/rubykov.rb
|
41
|
+
- lib/rubykov/hidden_markov_model.rb
|
41
42
|
- lib/rubykov/markov_model.rb
|
42
43
|
- lib/rubykov/text_generator.rb
|
44
|
+
- lib/rubykov/word_generator.rb
|
43
45
|
- license.md
|
44
46
|
homepage: https://gitlab.com/ehemsley/rubykov
|
45
47
|
licenses:
|
@@ -61,7 +63,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
61
63
|
version: '0'
|
62
64
|
requirements: []
|
63
65
|
rubyforge_project:
|
64
|
-
rubygems_version: 2.
|
66
|
+
rubygems_version: 2.5.1
|
65
67
|
signing_key:
|
66
68
|
specification_version: 4
|
67
69
|
summary: Sensible, easy Markov models in Ruby.
|