mhl 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/Gemfile +4 -0
- data/LICENSE +20 -0
- data/README.md +71 -0
- data/Rakefile +1 -0
- data/lib/mhl.rb +3 -0
- data/lib/mhl/bitstring_genotype_space.rb +84 -0
- data/lib/mhl/genetic_algorithm_solver.rb +164 -0
- data/lib/mhl/integer_genotype_space.rb +106 -0
- data/lib/mhl/particle_swarm_optimization_solver.rb +122 -0
- data/lib/mhl/version.rb +3 -0
- data/mhl.gemspec +28 -0
- data/spec/mhl/genetic_algorithm_spec.rb +53 -0
- data/spec/mhl/particle_swarm_optimization_solver_spec.rb +16 -0
- data/spec/spec_helper.rb +19 -0
- metadata +145 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: fd67377b3670dfd8f235044614d2f4b9965ead18
|
4
|
+
data.tar.gz: a3a98c4c0a8ead65030715cbf0e0f9f24f1d0a3c
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: a7dfe8e770e05c0bae580e3e0f21cb9da2accc690b05d19d6e1ce330dcae0e6a54ee3c95630809aabec34c9cc0af8a9befeb7ee1faabbbb87356e809522f2f46
|
7
|
+
data.tar.gz: cd3bb8ce03f86c22250d549251908d7ff2c7eb88b70a9108fe258213686afa7b9024eb0888607feb24031f912769e2073b6436c53ad7af059b90e78b462c8137
|
data/Gemfile
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
The MIT License (MIT)
|
2
|
+
|
3
|
+
Copyright (c) 2014 Mauro Tortonesi
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy of
|
6
|
+
this software and associated documentation files (the "Software"), to deal in
|
7
|
+
the Software without restriction, including without limitation the rights to
|
8
|
+
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
|
9
|
+
the Software, and to permit persons to whom the Software is furnished to do so,
|
10
|
+
subject to the following conditions:
|
11
|
+
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
13
|
+
copies or substantial portions of the Software.
|
14
|
+
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
|
17
|
+
FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
|
18
|
+
COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
|
19
|
+
IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
|
20
|
+
CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.md
ADDED
@@ -0,0 +1,71 @@
|
|
1
|
+
#ruby-mhl
|
2
|
+
|
3
|
+
A Ruby metaheuristics library
|
4
|
+
|
5
|
+
|
6
|
+
## Installation
|
7
|
+
|
8
|
+
### Stable version
|
9
|
+
|
10
|
+
You can get the stable version of ruby-mhl by installing the mhl gem from
|
11
|
+
RubyGems:
|
12
|
+
|
13
|
+
gem install mhl
|
14
|
+
|
15
|
+
### Development version
|
16
|
+
|
17
|
+
If you want to try the development version of ruby-mhl, instead, just place
|
18
|
+
this line:
|
19
|
+
|
20
|
+
```ruby
|
21
|
+
gem 'mhl', git: 'https://github.com/mtortonesi/ruby-mhl.git'
|
22
|
+
```
|
23
|
+
|
24
|
+
in your Gemfile and run:
|
25
|
+
|
26
|
+
bundle install
|
27
|
+
|
28
|
+
|
29
|
+
## Examples
|
30
|
+
|
31
|
+
Here is an example demonstrating how to find the argument that minimizes the
|
32
|
+
2-dimension parabola x_1 ^ 2 + x_2 ^ 2 equation with a genetic algorithm:
|
33
|
+
|
34
|
+
```ruby
|
35
|
+
require 'mhl'
|
36
|
+
|
37
|
+
solver = MHL::GeneticAlgorithmSolver.new(
|
38
|
+
:population_size => 40,
|
39
|
+
:genotype_space_type => :integer,
|
40
|
+
:mutation_probability => 0.5,
|
41
|
+
:recombination_probability => 0.5,
|
42
|
+
:genotype_space_conf => {
|
43
|
+
:dimensions => 2,
|
44
|
+
:recombination_type => :intermediate,
|
45
|
+
:random_func => lambda { Array.new(2) { rand(20) } }
|
46
|
+
},
|
47
|
+
:exit_condition => lambda {|generation,best| best[:fitness] == 0}
|
48
|
+
)
|
49
|
+
solver.solve(Proc.new{|x| -(x[0] ** 2 + x[1] ** 2) })
|
50
|
+
```
|
51
|
+
|
52
|
+
and with particle swarm optimization:
|
53
|
+
|
54
|
+
```ruby
|
55
|
+
require 'mhl'
|
56
|
+
|
57
|
+
solver = MHL::ParticleSwarmOptimizationSolver.new(
|
58
|
+
:swarm_size => 40,
|
59
|
+
:random_position_func => lambda { Array.new(2) { rand(20) } },
|
60
|
+
:random_velocity_func => lambda { Array.new(2) { rand(10) } },
|
61
|
+
:exit_condition => lambda {|generation,best| best[:height].abs < 0.001 },
|
62
|
+
)
|
63
|
+
solver.solve(Proc.new{|x| -(x[0] ** 2 + x[1] ** 2) })
|
64
|
+
```
|
65
|
+
|
66
|
+
Other examples and a full documentation will be publised as ruby-mhl matures.
|
67
|
+
|
68
|
+
|
69
|
+
## License
|
70
|
+
|
71
|
+
MIT
|
data/Rakefile
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
require 'bundler/gem_tasks'
|
data/lib/mhl.rb
ADDED
@@ -0,0 +1,84 @@
|
|
1
|
+
require 'bitstring'
|
2
|
+
|
3
|
+
module MHL
|
4
|
+
|
5
|
+
# This class implements a genotype with bitstring representation
|
6
|
+
class BitstringGenotypeSpace
|
7
|
+
def initialize(opts)
|
8
|
+
@bitstring_length = opts[:bitstring_length].to_i
|
9
|
+
unless @bitstring_length and @bitstring_length > 0
|
10
|
+
raise ArgumentError, 'Must have positive integer bitstring_length'
|
11
|
+
end
|
12
|
+
|
13
|
+
@random_func = opts[:random_func] || default_random_func(opts[:random_one_to_zero_ratio] || 1.0)
|
14
|
+
end
|
15
|
+
|
16
|
+
def get_random
|
17
|
+
@random_func.call
|
18
|
+
end
|
19
|
+
|
20
|
+
# reproduction with bitflip mutation and one-point crossover
|
21
|
+
def reproduce_from(p1, p2, mutation_rv, recombination_rv)
|
22
|
+
# make copies of p1 and p2
|
23
|
+
# (we're only interested in the :genotype key)
|
24
|
+
c1 = { :genotype => p1[:genotype].dup }
|
25
|
+
c2 = { :genotype => p2[:genotype].dup }
|
26
|
+
|
27
|
+
# mutation comes first
|
28
|
+
bitflip_mutation(c1[:genotype], mutation_rv)
|
29
|
+
bitflip_mutation(c2[:genotype], mutation_rv)
|
30
|
+
|
31
|
+
# and then recombination
|
32
|
+
c1[:genotype], c2[:genotype] =
|
33
|
+
onepoint_crossover(c1[:genotype], c2[:genotype], recombination_rv)
|
34
|
+
|
35
|
+
return c1, c2
|
36
|
+
end
|
37
|
+
|
38
|
+
private
|
39
|
+
|
40
|
+
def bitflip_mutation(bitstring, mutation_rv)
|
41
|
+
# TODO: disable this check in non-debugging mode
|
42
|
+
unless bitstring.length == @bitstring_length
|
43
|
+
raise 'Error! Different bit string sizes!'
|
44
|
+
end
|
45
|
+
|
46
|
+
@bitstring_length.times do |i|
|
47
|
+
if mutation_rv.next < @mutation_threshold
|
48
|
+
bitval = bitstring[i]
|
49
|
+
bitstring[i] = (bitval == 1 ? '0' : '1')
|
50
|
+
end
|
51
|
+
end
|
52
|
+
end
|
53
|
+
|
54
|
+
def onepoint_crossover(bitstring1, bitstring2, recombination_rv)
|
55
|
+
# TODO: disable this check in non-debugging mode
|
56
|
+
unless bitstring1.length == @bitstring_length and bitstring2.length == @bitstring_length
|
57
|
+
raise 'Error! Different bit string sizes!'
|
58
|
+
end
|
59
|
+
|
60
|
+
if recombination_rv.next < @recombination_threshold
|
61
|
+
size = bitstring1.length
|
62
|
+
point = 1 + rand(size - 2)
|
63
|
+
hi_mask = bitstring1.mask(point, BitString::LOW_END) # lowest point bits
|
64
|
+
low_mask = bitstring1.mask(size - point, BitString::HIGH_END) # highest size-point bits
|
65
|
+
new_b1 = (bitstring1 & hi_mask) | (bitstring2 & low_mask)
|
66
|
+
new_b2 = (bitstring2 & hi_mask) | (bitstring1 & low_mask)
|
67
|
+
return new_b1, new_b2
|
68
|
+
end
|
69
|
+
return bitstring1, bitstring2
|
70
|
+
end
|
71
|
+
|
72
|
+
def default_random_func(one_to_zero_ratio)
|
73
|
+
random_percentage_of_ones = one_to_zero_ratio / (1.0 + one_to_zero_ratio)
|
74
|
+
lambda do
|
75
|
+
str = (0...@bitstring_length).inject("") do |s,i|
|
76
|
+
s << ((rand < random_percentage_of_ones) ? '1' : '0')
|
77
|
+
end
|
78
|
+
BitString.new(str, size)
|
79
|
+
end
|
80
|
+
end
|
81
|
+
|
82
|
+
end
|
83
|
+
|
84
|
+
end
|
@@ -0,0 +1,164 @@
|
|
1
|
+
require 'concurrent'
|
2
|
+
require 'erv'
|
3
|
+
|
4
|
+
require 'mhl/bitstring_genotype_space'
|
5
|
+
require 'mhl/integer_genotype_space'
|
6
|
+
|
7
|
+
|
8
|
+
module MHL
|
9
|
+
|
10
|
+
class GeneticAlgorithmSolver
|
11
|
+
def initialize(opts)
|
12
|
+
@population_size = opts[:population_size].to_i
|
13
|
+
unless @population_size and @population_size.even?
|
14
|
+
raise ArgumentError, 'Even population size required!'
|
15
|
+
end
|
16
|
+
|
17
|
+
# perform genotype space-specific configuration
|
18
|
+
case opts[:genotype_space_type]
|
19
|
+
when :integer
|
20
|
+
@genotype_space = IntegerVectorGenotypeSpace.new(opts[:genotype_space_conf])
|
21
|
+
|
22
|
+
begin
|
23
|
+
p_m = opts[:mutation_probability].to_f
|
24
|
+
@mutation_rv = \
|
25
|
+
ERV::RandomVariable.new(:distribution => :geometric,
|
26
|
+
:probability_of_success => p_m)
|
27
|
+
rescue
|
28
|
+
raise ArgumentError, 'Mutation probability configuration is wrong.'
|
29
|
+
end
|
30
|
+
|
31
|
+
begin
|
32
|
+
p_r = opts[:recombination_probability].to_f
|
33
|
+
@recombination_rv = \
|
34
|
+
ERV::RandomVariable.new(:distribution => :uniform,
|
35
|
+
:min_value => -p_r,
|
36
|
+
:max_value => 1.0 + p_r)
|
37
|
+
rescue
|
38
|
+
raise ArgumentError, 'Recombination probability configuration is wrong.'
|
39
|
+
end
|
40
|
+
|
41
|
+
when :bitstring
|
42
|
+
@genotype_space = BitstringGenotypeSpace.new(opts[:genotype_space_conf])
|
43
|
+
@recombination_rv = ERV::RandomVariable.new(:distribution => :uniform, :max_value => 1.0)
|
44
|
+
@mutation_rv = ERV::RandomVariable.new(:distribution => :uniform, :max_value => 1.0)
|
45
|
+
|
46
|
+
else
|
47
|
+
raise ArgumentError, 'Only integer and bitstring genotype representations are supported!'
|
48
|
+
end
|
49
|
+
|
50
|
+
@exit_condition = opts[:exit_condition]
|
51
|
+
@start_population = opts[:genotype_space_conf][:start_population]
|
52
|
+
end
|
53
|
+
|
54
|
+
|
55
|
+
# This is the method that solves the optimization problem
|
56
|
+
#
|
57
|
+
# Parameter func is supposed to be a method (or a Proc, a lambda, or any callable
|
58
|
+
# object) that accepts the genotype as argument (that is, the set of
|
59
|
+
# parameters) and returns the phenotype (that is, the function result)
|
60
|
+
def solve(func)
|
61
|
+
# setup population
|
62
|
+
if @start_population.nil?
|
63
|
+
population = Array.new(@population_size) do
|
64
|
+
# generate random genotype according to the chromosome type
|
65
|
+
{ :genotype => @genotype_space.get_random }
|
66
|
+
end
|
67
|
+
else
|
68
|
+
population = @start_population.map do |x|
|
69
|
+
{ :genotype => x }
|
70
|
+
end
|
71
|
+
end
|
72
|
+
|
73
|
+
# initialize variables
|
74
|
+
gen = 0
|
75
|
+
overall_best = nil
|
76
|
+
|
77
|
+
# default behavior is to loop forever
|
78
|
+
begin
|
79
|
+
gen += 1
|
80
|
+
puts "Starting generation #{gen} at #{Time.now}"
|
81
|
+
|
82
|
+
# assess fitness for every member of the population
|
83
|
+
population.each do |s|
|
84
|
+
s[:task] = Concurrent::Future.new { func.call(s[:genotype]) }
|
85
|
+
end
|
86
|
+
|
87
|
+
# wait for all the evaluations to end
|
88
|
+
population.each do |s|
|
89
|
+
s[:fitness] = s[:task].value
|
90
|
+
end
|
91
|
+
|
92
|
+
# find fittest member
|
93
|
+
population_best = population.max_by {|x| x[:fitness] }
|
94
|
+
|
95
|
+
# calculate overall best
|
96
|
+
if overall_best.nil?
|
97
|
+
overall_best = population_best
|
98
|
+
else
|
99
|
+
overall_best = [ overall_best, population_best ].max_by {|x| x[:fitness] }
|
100
|
+
end
|
101
|
+
|
102
|
+
# print results
|
103
|
+
puts "> gen #{gen}, best: #{overall_best[:genotype]}, #{overall_best[:fitness]}"
|
104
|
+
|
105
|
+
# selection by binary tournament
|
106
|
+
children = new_generation(population)
|
107
|
+
|
108
|
+
# update population and generation number
|
109
|
+
population = children
|
110
|
+
end while @exit_condition.nil? or !@exit_condition.call(gen, overall_best)
|
111
|
+
end
|
112
|
+
|
113
|
+
|
114
|
+
private
|
115
|
+
|
116
|
+
# reproduction with point mutation and one-point crossover
|
117
|
+
def new_generation(population)
|
118
|
+
population_size = population.size
|
119
|
+
|
120
|
+
# check correct population size
|
121
|
+
# TODO: disable this check in non-debugging mode
|
122
|
+
raise ArgumentError, 'Population size error!' if population_size != @population_size
|
123
|
+
|
124
|
+
# prepare children
|
125
|
+
children = []
|
126
|
+
|
127
|
+
# select members to reproduce through binary tournament
|
128
|
+
selected = Array.new(@population_size) { |i| binary_tournament(population) }
|
129
|
+
selected.shuffle!
|
130
|
+
|
131
|
+
# reproduction
|
132
|
+
selected.each_slice(2) do |p1, p2|
|
133
|
+
# get two new samples...
|
134
|
+
c1, c2 = @genotype_space.reproduce_from(p1, p2, @mutation_rv, @recombination_rv)
|
135
|
+
|
136
|
+
# ...and add them to the children population
|
137
|
+
children.push(c1, c2)
|
138
|
+
|
139
|
+
# check correct population size
|
140
|
+
# TODO: disable this check in non-debugging mode
|
141
|
+
raise 'Children size error!' if children.size > population_size
|
142
|
+
end
|
143
|
+
|
144
|
+
return children
|
145
|
+
end
|
146
|
+
|
147
|
+
# This method implements binary tournament selection, which is probably
|
148
|
+
# the most popular selection method for genetic algorithms
|
149
|
+
def binary_tournament(population)
|
150
|
+
i = rand(population.size)
|
151
|
+
j = rand(population.size - 1)
|
152
|
+
j += 1 if j >= i
|
153
|
+
|
154
|
+
select_fittest(population[i], population[j])
|
155
|
+
end
|
156
|
+
|
157
|
+
def select_fittest(*a)
|
158
|
+
# TODO: disable this check in non-debugging mode
|
159
|
+
raise 'Attempting to select the fittest sample of an empty population!' if a.empty?
|
160
|
+
a.max_by {|x| x[:fitness] }
|
161
|
+
end
|
162
|
+
end
|
163
|
+
|
164
|
+
end
|
@@ -0,0 +1,106 @@
|
|
1
|
+
module MHL
|
2
|
+
|
3
|
+
# This class implements a genotype with integer space representation
|
4
|
+
class IntegerVectorGenotypeSpace
|
5
|
+
def initialize(opts)
|
6
|
+
@random_func = opts[:random_func]
|
7
|
+
|
8
|
+
@dimensions = opts[:dimensions].to_i
|
9
|
+
unless @dimensions and @dimensions > 0
|
10
|
+
raise ArgumentError, 'Must have positive integer dimensions'
|
11
|
+
end
|
12
|
+
|
13
|
+
# TODO: enable to choose which recombination function to use
|
14
|
+
case opts[:recombination_type].to_s
|
15
|
+
when /intermediate/i
|
16
|
+
@recombination_func = :intermediate_recombination
|
17
|
+
when /line/i
|
18
|
+
@recombination_func = :line_recombination
|
19
|
+
else
|
20
|
+
raise ArgumentError, 'Recombination function must be either line or intermediate!'
|
21
|
+
end
|
22
|
+
end
|
23
|
+
|
24
|
+
def get_random
|
25
|
+
if @random_func
|
26
|
+
@random_func.call
|
27
|
+
else
|
28
|
+
# TODO: implement this
|
29
|
+
end
|
30
|
+
end
|
31
|
+
|
32
|
+
# reproduction with random geometric mutation
|
33
|
+
# and intermediate recombination
|
34
|
+
def reproduce_from(p1, p2, mutation_rv, recombination_rv)
|
35
|
+
# make copies of p1 and p2
|
36
|
+
# (we're only interested in the :genotype key)
|
37
|
+
c1 = { :genotype => p1[:genotype].dup }
|
38
|
+
c2 = { :genotype => p2[:genotype].dup }
|
39
|
+
|
40
|
+
# mutation comes first
|
41
|
+
random_geometric_mutation(c1[:genotype], mutation_rv)
|
42
|
+
random_geometric_mutation(c2[:genotype], mutation_rv)
|
43
|
+
|
44
|
+
# and then recombination
|
45
|
+
send(@recombination_func, c1[:genotype], c2[:genotype], recombination_rv)
|
46
|
+
|
47
|
+
return c1, c2
|
48
|
+
end
|
49
|
+
|
50
|
+
|
51
|
+
private
|
52
|
+
|
53
|
+
def random_geometric_mutation(g, mutation_rv)
|
54
|
+
g.each_index do |i|
|
55
|
+
# being sampled from a geometric distribution, delta will always
|
56
|
+
# be a non-negative integer (that is, 0 or greater)
|
57
|
+
delta = mutation_rv.next
|
58
|
+
|
59
|
+
if rand() >= 0.5
|
60
|
+
# half of the times the variation will be positive ...
|
61
|
+
g[i] += delta
|
62
|
+
else
|
63
|
+
# ... and half of the times it will be negative
|
64
|
+
g[i] -= delta
|
65
|
+
end
|
66
|
+
end
|
67
|
+
end
|
68
|
+
|
69
|
+
def intermediate_recombination(g1, g2, recombination_rv)
|
70
|
+
# TODO: disable this check in non-debugging mode
|
71
|
+
raise ArgumentError, 'g1 and g2 must have the same dimension' unless g1.size == g2.size
|
72
|
+
|
73
|
+
# recombination
|
74
|
+
g1.each_index do |i|
|
75
|
+
begin
|
76
|
+
alpha = recombination_rv.next
|
77
|
+
beta = recombination_rv.next
|
78
|
+
t = (alpha * g1[i] + (1.0 - alpha) * g2[i] + 0.5).floor
|
79
|
+
s = ( beta * g2[i] + (1.0 - beta) * g1[i] + 0.5).floor
|
80
|
+
end # until t >= 0 and s >= 0 # TODO: implement within-bounds condition checking
|
81
|
+
g1[i] = t
|
82
|
+
g2[i] = s
|
83
|
+
end
|
84
|
+
end
|
85
|
+
|
86
|
+
def line_recombination(g1, g2, recombination_rv)
|
87
|
+
# TODO: disable this check in non-debugging mode
|
88
|
+
raise ArgumentError, 'g1 and g2 must have the same dimension' unless g1.size == g2.size
|
89
|
+
|
90
|
+
alpha = recombination_rv.next
|
91
|
+
beta = recombination_rv.next
|
92
|
+
|
93
|
+
# recombination
|
94
|
+
g1.each_index do |i|
|
95
|
+
t = (alpha * g1[i] + (1.0 - alpha) * g2[i] + 0.5).floor
|
96
|
+
s = ( beta * g2[i] + (1.0 - beta) * g1[i] + 0.5).floor
|
97
|
+
# if t >= 0 and s >= 0 # TODO: implement within-bounds condition checking
|
98
|
+
g1[i] = t
|
99
|
+
g2[i] = s
|
100
|
+
# end
|
101
|
+
end
|
102
|
+
end
|
103
|
+
|
104
|
+
end
|
105
|
+
|
106
|
+
end
|
@@ -0,0 +1,122 @@
|
|
1
|
+
require 'matrix'
|
2
|
+
require 'securerandom'
|
3
|
+
|
4
|
+
module MHL
|
5
|
+
|
6
|
+
class ParticleSwarmOptimizationSolver
|
7
|
+
|
8
|
+
def initialize(opts={})
|
9
|
+
@swarm_size = opts[:swarm_size].to_i
|
10
|
+
unless @swarm_size
|
11
|
+
raise ArgumentError, 'Swarm size is a required parameter!'
|
12
|
+
end
|
13
|
+
|
14
|
+
@random_position_func = opts[:random_position_func]
|
15
|
+
@random_velocity_func = opts[:random_velocity_func]
|
16
|
+
|
17
|
+
@start_positions = opts[:start_positions]
|
18
|
+
@exit_condition = opts[:exit_condition]
|
19
|
+
end
|
20
|
+
|
21
|
+
# This is the method that solves the optimization problem
|
22
|
+
#
|
23
|
+
# Parameter func is supposed to be a method (or a Proc, a lambda, or any callable
|
24
|
+
# object) that accepts the genotype as argument (that is, the set of
|
25
|
+
# parameters) and returns the phenotype (that is, the function result)
|
26
|
+
def solve(func)
|
27
|
+
# setup particles
|
28
|
+
if @start_positions.nil?
|
29
|
+
particles = Array.new(@swarm_size) do
|
30
|
+
{ position: Vector[*@random_position_func.call], velocity: Vector[*@random_velocity_func.call] }
|
31
|
+
end
|
32
|
+
else
|
33
|
+
particles = @start_positions.each_slice(2).map do |pos,vel|
|
34
|
+
{ position: Vector[*pos], velocity: Vector[*vel] }
|
35
|
+
end
|
36
|
+
end
|
37
|
+
|
38
|
+
# initialize variables
|
39
|
+
gen = 0
|
40
|
+
overall_best = nil
|
41
|
+
|
42
|
+
# completely made up values
|
43
|
+
alpha = 0.5
|
44
|
+
beta = 0.3
|
45
|
+
gamma = 0.7
|
46
|
+
delta = 0.5
|
47
|
+
epsilon = 0.6
|
48
|
+
|
49
|
+
# default behavior is to loop forever
|
50
|
+
begin
|
51
|
+
gen += 1
|
52
|
+
puts "Starting generation #{gen} at #{Time.now}"
|
53
|
+
|
54
|
+
# assess height for every particle
|
55
|
+
particles.each do |p|
|
56
|
+
p[:task] = Concurrent::Future.new { func.call(p[:position]) }
|
57
|
+
end
|
58
|
+
|
59
|
+
# wait for all the evaluations to end
|
60
|
+
particles.each_with_index do |p,i|
|
61
|
+
p[:height] = p[:task].value
|
62
|
+
if p[:highest_value].nil? or p[:height] > p[:highest_value]
|
63
|
+
p[:highest_value] = p[:height]
|
64
|
+
p[:highest_position] = p[:position]
|
65
|
+
end
|
66
|
+
end
|
67
|
+
|
68
|
+
# find highest particle
|
69
|
+
highest_particle = particles.max_by {|x| x[:height] }
|
70
|
+
|
71
|
+
# calculate overall best
|
72
|
+
if overall_best.nil?
|
73
|
+
overall_best = highest_particle
|
74
|
+
else
|
75
|
+
overall_best = [ overall_best, highest_particle ].max_by {|x| x[:height] }
|
76
|
+
end
|
77
|
+
|
78
|
+
# mutate swarm
|
79
|
+
particles.each do |p|
|
80
|
+
# randomly sample particles and use them as informants
|
81
|
+
informants = random_portion(particles)
|
82
|
+
|
83
|
+
# make sure that p is included among the informants
|
84
|
+
informants << p unless informants.include? p
|
85
|
+
|
86
|
+
# get fittest informant
|
87
|
+
fittest_informant = informants.max_by {|x| x[:height] }
|
88
|
+
|
89
|
+
# update velocity
|
90
|
+
p[:velocity] =
|
91
|
+
alpha * p[:velocity] +
|
92
|
+
beta * (p[:highest_position] - p[:position]) +
|
93
|
+
gamma * (fittest_informant[:highest_position] - p[:position]) +
|
94
|
+
delta * (overall_best[:highest_position] - p[:position])
|
95
|
+
|
96
|
+
# update position
|
97
|
+
p[:position] = p[:position] + epsilon * p[:velocity]
|
98
|
+
end
|
99
|
+
|
100
|
+
end while @exit_condition.nil? or !@exit_condition.call(gen, overall_best)
|
101
|
+
end
|
102
|
+
|
103
|
+
private
|
104
|
+
|
105
|
+
def random_portion(array, ratio=0.1)
|
106
|
+
# get size of random array to return
|
107
|
+
size = (ratio * array.size).ceil
|
108
|
+
|
109
|
+
(1..size).inject([]) do |acc,i|
|
110
|
+
# randomly sample a new element
|
111
|
+
begin
|
112
|
+
new_element = array[SecureRandom.random_number(array.size)]
|
113
|
+
end while acc.include? new_element
|
114
|
+
|
115
|
+
# insert element in the accumulator
|
116
|
+
acc << new_element
|
117
|
+
end
|
118
|
+
end
|
119
|
+
|
120
|
+
end
|
121
|
+
|
122
|
+
end
|
data/lib/mhl/version.rb
ADDED
data/mhl.gemspec
ADDED
@@ -0,0 +1,28 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
lib = File.expand_path('../lib', __FILE__)
|
3
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
4
|
+
require 'mhl/version'
|
5
|
+
|
6
|
+
Gem::Specification.new do |spec|
|
7
|
+
spec.name = 'mhl'
|
8
|
+
spec.version = MHL::VERSION
|
9
|
+
spec.authors = ['Mauro Tortonesi']
|
10
|
+
spec.email = ['mauro.tortonesi@unife.it']
|
11
|
+
spec.description = %q{A Ruby Metaheuristics library}
|
12
|
+
spec.summary = %q{A scientific library for Ruby that provides several metaheuristics}
|
13
|
+
spec.homepage = 'https://github.com/mtortonesi/ruby-mhl'
|
14
|
+
spec.license = 'MIT'
|
15
|
+
|
16
|
+
spec.files = `git ls-files`.split($/).reject{|x| x == '.gitignore' }
|
17
|
+
spec.executables = spec.files.grep(%r{^bin/}) { |f| File.basename(f) }
|
18
|
+
spec.test_files = spec.files.grep(%r{^(test|spec|features)/})
|
19
|
+
spec.require_paths = ['lib']
|
20
|
+
|
21
|
+
spec.add_dependency 'bitstring'
|
22
|
+
spec.add_dependency 'concurrent-ruby'
|
23
|
+
spec.add_dependency 'erv'
|
24
|
+
|
25
|
+
spec.add_development_dependency 'bundler', '~> 1.3'
|
26
|
+
spec.add_development_dependency 'rake'
|
27
|
+
spec.add_development_dependency 'rspec'
|
28
|
+
end
|
@@ -0,0 +1,53 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe MHL::GeneticAlgorithmSolver do
|
4
|
+
|
5
|
+
it 'should accept bitstring representation genotypes' do
|
6
|
+
lambda {
|
7
|
+
MHL::GeneticAlgorithmSolver.new(
|
8
|
+
:population_size => 128,
|
9
|
+
:genotype_space_type => :bitstring,
|
10
|
+
:mutation_threshold => 0.5,
|
11
|
+
:recombination_threshold => 0.5,
|
12
|
+
:genotype_space_conf => {
|
13
|
+
:bitstring_length => 120,
|
14
|
+
}
|
15
|
+
)
|
16
|
+
}.should_not raise_error
|
17
|
+
end
|
18
|
+
|
19
|
+
it 'should accept integer representation genotypes' do
|
20
|
+
lambda {
|
21
|
+
MHL::GeneticAlgorithmSolver.new(
|
22
|
+
:population_size => 128,
|
23
|
+
:genotype_space_type => :integer,
|
24
|
+
:mutation_probability => 0.5,
|
25
|
+
:recombination_probability => 0.5,
|
26
|
+
:genotype_space_conf => {
|
27
|
+
:dimensions => 6,
|
28
|
+
:recombination_type => :intermediate,
|
29
|
+
}
|
30
|
+
)
|
31
|
+
}.should_not raise_error
|
32
|
+
end
|
33
|
+
|
34
|
+
it 'should solve a 2-dimension parabola in integer space' do
|
35
|
+
solver = MHL::GeneticAlgorithmSolver.new(
|
36
|
+
:population_size => 40,
|
37
|
+
:genotype_space_type => :integer,
|
38
|
+
:mutation_probability => 0.5,
|
39
|
+
:recombination_probability => 0.5,
|
40
|
+
:genotype_space_conf => {
|
41
|
+
:dimensions => 2,
|
42
|
+
:recombination_type => :intermediate,
|
43
|
+
:random_func => lambda { Array.new(2) { rand(20) } }
|
44
|
+
},
|
45
|
+
:exit_condition => lambda {|generation,best_sample| best_sample[:fitness] == 0}
|
46
|
+
)
|
47
|
+
solver.solve(Proc.new{|genotype| -(genotype[0]**2 + genotype[1]**2) })
|
48
|
+
end
|
49
|
+
|
50
|
+
# it 'should solve a 2-dimension parabola in real space'
|
51
|
+
|
52
|
+
end
|
53
|
+
|
@@ -0,0 +1,16 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe MHL::ParticleSwarmOptimizationSolver do
|
4
|
+
|
5
|
+
it 'should solve a 2-dimension parabola in real space' do
|
6
|
+
solver = MHL::ParticleSwarmOptimizationSolver.new(
|
7
|
+
:swarm_size => 40,
|
8
|
+
:random_position_func => lambda { Array.new(2) { rand(20) } },
|
9
|
+
:random_velocity_func => lambda { Array.new(2) { rand(10) } },
|
10
|
+
:exit_condition => lambda {|generation,best_sample| best_sample[:height].abs < 0.001 },
|
11
|
+
)
|
12
|
+
solver.solve(Proc.new{|position| -(position[0]**2 + position[1]**2) })
|
13
|
+
end
|
14
|
+
|
15
|
+
end
|
16
|
+
|
data/spec/spec_helper.rb
ADDED
@@ -0,0 +1,19 @@
|
|
1
|
+
# This file was generated by the `rspec --init` command. Conventionally, all
|
2
|
+
# specs live under a `spec` directory, which RSpec adds to the `$LOAD_PATH`.
|
3
|
+
# Require this file using `require "spec_helper"` to ensure that it is only
|
4
|
+
# loaded once.
|
5
|
+
|
6
|
+
require 'mhl'
|
7
|
+
|
8
|
+
# See http://rubydoc.info/gems/rspec-core/RSpec/Core/Configuration
|
9
|
+
RSpec.configure do |config|
|
10
|
+
config.treat_symbols_as_metadata_keys_with_true_values = true
|
11
|
+
config.run_all_when_everything_filtered = true
|
12
|
+
config.filter_run :focus
|
13
|
+
|
14
|
+
# Run specs in random order to surface order dependencies. If you find an
|
15
|
+
# order dependency and want to debug it, you can fix the order by providing
|
16
|
+
# the seed, which is printed after each run.
|
17
|
+
# --seed 1234
|
18
|
+
config.order = 'random'
|
19
|
+
end
|
metadata
ADDED
@@ -0,0 +1,145 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: mhl
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Mauro Tortonesi
|
8
|
+
autorequire:
|
9
|
+
bindir: bin
|
10
|
+
cert_chain: []
|
11
|
+
date: 2014-02-06 00:00:00.000000000 Z
|
12
|
+
dependencies:
|
13
|
+
- !ruby/object:Gem::Dependency
|
14
|
+
name: bitstring
|
15
|
+
version_requirements: !ruby/object:Gem::Requirement
|
16
|
+
requirements:
|
17
|
+
- - '>='
|
18
|
+
- !ruby/object:Gem::Version
|
19
|
+
version: '0'
|
20
|
+
requirement: !ruby/object:Gem::Requirement
|
21
|
+
requirements:
|
22
|
+
- - '>='
|
23
|
+
- !ruby/object:Gem::Version
|
24
|
+
version: '0'
|
25
|
+
prerelease: false
|
26
|
+
type: :runtime
|
27
|
+
- !ruby/object:Gem::Dependency
|
28
|
+
name: concurrent-ruby
|
29
|
+
version_requirements: !ruby/object:Gem::Requirement
|
30
|
+
requirements:
|
31
|
+
- - '>='
|
32
|
+
- !ruby/object:Gem::Version
|
33
|
+
version: '0'
|
34
|
+
requirement: !ruby/object:Gem::Requirement
|
35
|
+
requirements:
|
36
|
+
- - '>='
|
37
|
+
- !ruby/object:Gem::Version
|
38
|
+
version: '0'
|
39
|
+
prerelease: false
|
40
|
+
type: :runtime
|
41
|
+
- !ruby/object:Gem::Dependency
|
42
|
+
name: erv
|
43
|
+
version_requirements: !ruby/object:Gem::Requirement
|
44
|
+
requirements:
|
45
|
+
- - '>='
|
46
|
+
- !ruby/object:Gem::Version
|
47
|
+
version: '0'
|
48
|
+
requirement: !ruby/object:Gem::Requirement
|
49
|
+
requirements:
|
50
|
+
- - '>='
|
51
|
+
- !ruby/object:Gem::Version
|
52
|
+
version: '0'
|
53
|
+
prerelease: false
|
54
|
+
type: :runtime
|
55
|
+
- !ruby/object:Gem::Dependency
|
56
|
+
name: bundler
|
57
|
+
version_requirements: !ruby/object:Gem::Requirement
|
58
|
+
requirements:
|
59
|
+
- - ~>
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: '1.3'
|
62
|
+
requirement: !ruby/object:Gem::Requirement
|
63
|
+
requirements:
|
64
|
+
- - ~>
|
65
|
+
- !ruby/object:Gem::Version
|
66
|
+
version: '1.3'
|
67
|
+
prerelease: false
|
68
|
+
type: :development
|
69
|
+
- !ruby/object:Gem::Dependency
|
70
|
+
name: rake
|
71
|
+
version_requirements: !ruby/object:Gem::Requirement
|
72
|
+
requirements:
|
73
|
+
- - '>='
|
74
|
+
- !ruby/object:Gem::Version
|
75
|
+
version: '0'
|
76
|
+
requirement: !ruby/object:Gem::Requirement
|
77
|
+
requirements:
|
78
|
+
- - '>='
|
79
|
+
- !ruby/object:Gem::Version
|
80
|
+
version: '0'
|
81
|
+
prerelease: false
|
82
|
+
type: :development
|
83
|
+
- !ruby/object:Gem::Dependency
|
84
|
+
name: rspec
|
85
|
+
version_requirements: !ruby/object:Gem::Requirement
|
86
|
+
requirements:
|
87
|
+
- - '>='
|
88
|
+
- !ruby/object:Gem::Version
|
89
|
+
version: '0'
|
90
|
+
requirement: !ruby/object:Gem::Requirement
|
91
|
+
requirements:
|
92
|
+
- - '>='
|
93
|
+
- !ruby/object:Gem::Version
|
94
|
+
version: '0'
|
95
|
+
prerelease: false
|
96
|
+
type: :development
|
97
|
+
description: A Ruby Metaheuristics library
|
98
|
+
email:
|
99
|
+
- mauro.tortonesi@unife.it
|
100
|
+
executables: []
|
101
|
+
extensions: []
|
102
|
+
extra_rdoc_files: []
|
103
|
+
files:
|
104
|
+
- Gemfile
|
105
|
+
- LICENSE
|
106
|
+
- README.md
|
107
|
+
- Rakefile
|
108
|
+
- lib/mhl.rb
|
109
|
+
- lib/mhl/bitstring_genotype_space.rb
|
110
|
+
- lib/mhl/genetic_algorithm_solver.rb
|
111
|
+
- lib/mhl/integer_genotype_space.rb
|
112
|
+
- lib/mhl/particle_swarm_optimization_solver.rb
|
113
|
+
- lib/mhl/version.rb
|
114
|
+
- mhl.gemspec
|
115
|
+
- spec/mhl/genetic_algorithm_spec.rb
|
116
|
+
- spec/mhl/particle_swarm_optimization_solver_spec.rb
|
117
|
+
- spec/spec_helper.rb
|
118
|
+
homepage: https://github.com/mtortonesi/ruby-mhl
|
119
|
+
licenses:
|
120
|
+
- MIT
|
121
|
+
metadata: {}
|
122
|
+
post_install_message:
|
123
|
+
rdoc_options: []
|
124
|
+
require_paths:
|
125
|
+
- lib
|
126
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
127
|
+
requirements:
|
128
|
+
- - '>='
|
129
|
+
- !ruby/object:Gem::Version
|
130
|
+
version: '0'
|
131
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
132
|
+
requirements:
|
133
|
+
- - '>='
|
134
|
+
- !ruby/object:Gem::Version
|
135
|
+
version: '0'
|
136
|
+
requirements: []
|
137
|
+
rubyforge_project:
|
138
|
+
rubygems_version: 2.2.1
|
139
|
+
signing_key:
|
140
|
+
specification_version: 4
|
141
|
+
summary: A scientific library for Ruby that provides several metaheuristics
|
142
|
+
test_files:
|
143
|
+
- spec/mhl/genetic_algorithm_spec.rb
|
144
|
+
- spec/mhl/particle_swarm_optimization_solver_spec.rb
|
145
|
+
- spec/spec_helper.rb
|