affinity_propagation 0.1.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +54 -0
- data/.rspec +2 -0
- data/.ruby-gemset +1 -0
- data/.ruby-version +1 -0
- data/.travis.yml +5 -0
- data/Gemfile +6 -0
- data/Gemfile.lock +47 -0
- data/LICENSE.txt +21 -0
- data/README.md +121 -0
- data/Rakefile +6 -0
- data/affinity_propagation.gemspec +32 -0
- data/bin/console +14 -0
- data/bin/setup +8 -0
- data/lib/affinity_propagation.rb +6 -0
- data/lib/affinity_propagation/calculator.rb +180 -0
- data/lib/affinity_propagation/version.rb +3 -0
- metadata +121 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: 26f8c5ca0ecef4d7c7e89341a5039eadefef5871
|
4
|
+
data.tar.gz: cd466acdd9604ca6e547558438203b88283d21c9
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 72c71cf377b691208ffeb714022216bb3d4b259ec09157feded788f49858d8e5260538ab73dec78e1a126b9914c9eb8c251a0c44818235ba35c366b81f24cd28
|
7
|
+
data.tar.gz: 4fdcd1c0742094f837c95ebcd4366045006deb97dd358fe55a6575244ad07eb9207a3b5331cf9f8035fb2ec9dca2ad0b46d68188eb23d79b0a87646e79542ab5
|
data/.gitignore
ADDED
@@ -0,0 +1,54 @@
|
|
1
|
+
*.gem
|
2
|
+
*.rbc
|
3
|
+
/.config
|
4
|
+
/coverage/
|
5
|
+
/InstalledFiles
|
6
|
+
/pkg/
|
7
|
+
/spec/reports/
|
8
|
+
/spec/examples.txt
|
9
|
+
/test/tmp/
|
10
|
+
/test/version_tmp/
|
11
|
+
/tmp/
|
12
|
+
|
13
|
+
# Used by dotenv library to load environment variables.
|
14
|
+
# .env
|
15
|
+
|
16
|
+
## Specific to RubyMotion:
|
17
|
+
.dat*
|
18
|
+
.repl_history
|
19
|
+
build/
|
20
|
+
*.bridgesupport
|
21
|
+
build-iPhoneOS/
|
22
|
+
build-iPhoneSimulator/
|
23
|
+
|
24
|
+
## Specific to RubyMotion (use of CocoaPods):
|
25
|
+
#
|
26
|
+
# We recommend against adding the Pods directory to your .gitignore. However
|
27
|
+
# you should judge for yourself, the pros and cons are mentioned at:
|
28
|
+
# https://guides.cocoapods.org/using/using-cocoapods.html#should-i-check-the-pods-directory-into-source-control
|
29
|
+
#
|
30
|
+
# vendor/Pods/
|
31
|
+
|
32
|
+
## Documentation cache and generated files:
|
33
|
+
/.yardoc/
|
34
|
+
/_yardoc/
|
35
|
+
/doc/
|
36
|
+
/rdoc/
|
37
|
+
|
38
|
+
## Environment normalization:
|
39
|
+
/.bundle/
|
40
|
+
/vendor/bundle
|
41
|
+
/lib/bundler/man/
|
42
|
+
|
43
|
+
# for a library or gem, you might want to ignore these files since the code is
|
44
|
+
# intended to run in multiple environments; otherwise, check them in:
|
45
|
+
# Gemfile.lock
|
46
|
+
# .ruby-version
|
47
|
+
# .ruby-gemset
|
48
|
+
|
49
|
+
# unless supporting rvm < 1.11.0 or doing something fancy, ignore this:
|
50
|
+
.rvmrc
|
51
|
+
|
52
|
+
# IDEs
|
53
|
+
.idea
|
54
|
+
|
data/.rspec
ADDED
data/.ruby-gemset
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
affinity_propagation
|
data/.ruby-version
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
ruby-2.4.1
|
data/.travis.yml
ADDED
data/Gemfile
ADDED
data/Gemfile.lock
ADDED
@@ -0,0 +1,47 @@
|
|
1
|
+
PATH
|
2
|
+
remote: .
|
3
|
+
specs:
|
4
|
+
affinity_propagation (0.1.0)
|
5
|
+
|
6
|
+
GEM
|
7
|
+
remote: https://rubygems.org/
|
8
|
+
specs:
|
9
|
+
activesupport (5.1.4)
|
10
|
+
concurrent-ruby (~> 1.0, >= 1.0.2)
|
11
|
+
i18n (~> 0.7)
|
12
|
+
minitest (~> 5.1)
|
13
|
+
tzinfo (~> 1.1)
|
14
|
+
concurrent-ruby (1.0.5)
|
15
|
+
diff-lcs (1.3)
|
16
|
+
i18n (0.8.6)
|
17
|
+
minitest (5.10.3)
|
18
|
+
rake (10.5.0)
|
19
|
+
rspec (3.6.0)
|
20
|
+
rspec-core (~> 3.6.0)
|
21
|
+
rspec-expectations (~> 3.6.0)
|
22
|
+
rspec-mocks (~> 3.6.0)
|
23
|
+
rspec-core (3.6.0)
|
24
|
+
rspec-support (~> 3.6.0)
|
25
|
+
rspec-expectations (3.6.0)
|
26
|
+
diff-lcs (>= 1.2.0, < 2.0)
|
27
|
+
rspec-support (~> 3.6.0)
|
28
|
+
rspec-mocks (3.6.0)
|
29
|
+
diff-lcs (>= 1.2.0, < 2.0)
|
30
|
+
rspec-support (~> 3.6.0)
|
31
|
+
rspec-support (3.6.0)
|
32
|
+
thread_safe (0.3.6)
|
33
|
+
tzinfo (1.2.3)
|
34
|
+
thread_safe (~> 0.1)
|
35
|
+
|
36
|
+
PLATFORMS
|
37
|
+
ruby
|
38
|
+
|
39
|
+
DEPENDENCIES
|
40
|
+
activesupport
|
41
|
+
affinity_propagation!
|
42
|
+
bundler (~> 1.15)
|
43
|
+
rake (~> 10.0)
|
44
|
+
rspec (~> 3.0)
|
45
|
+
|
46
|
+
BUNDLED WITH
|
47
|
+
1.15.4
|
data/LICENSE.txt
ADDED
@@ -0,0 +1,21 @@
|
|
1
|
+
The MIT License (MIT)
|
2
|
+
|
3
|
+
Copyright (c) 2017 Shahbaz Javeed
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
7
|
+
in the Software without restriction, including without limitation the rights
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
10
|
+
furnished to do so, subject to the following conditions:
|
11
|
+
|
12
|
+
The above copyright notice and this permission notice shall be included in
|
13
|
+
all 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,
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
21
|
+
THE SOFTWARE.
|
data/README.md
ADDED
@@ -0,0 +1,121 @@
|
|
1
|
+
# AffinityPropagation
|
2
|
+
|
3
|
+
This gem allows you to use the affinity propagation algorithm (see BJ Frey and D Dueck,
|
4
|
+
[Science 315, Feb 16, 2007](http://www.psi.toronto.edu/affinitypropagation/FreyDueckScience07.pdf)) to cluster arbitrary
|
5
|
+
data by providing the data and a block to calculate the similarity between any two data points in that list.
|
6
|
+
|
7
|
+
For more information on affinity propagation, visit the [FAQ over at University of Toronto](http://www.psi.toronto.edu/affinitypropagation/faq.html).
|
8
|
+
|
9
|
+
## Installation
|
10
|
+
|
11
|
+
Add this line to your application's Gemfile:
|
12
|
+
|
13
|
+
```ruby
|
14
|
+
gem 'affinity_propagation'
|
15
|
+
```
|
16
|
+
|
17
|
+
And then execute:
|
18
|
+
|
19
|
+
$ bundle
|
20
|
+
|
21
|
+
Or install it yourself as:
|
22
|
+
|
23
|
+
$ gem install affinity_propagation
|
24
|
+
|
25
|
+
## Usage
|
26
|
+
|
27
|
+
A simple example of usage is:
|
28
|
+
|
29
|
+
```ruby
|
30
|
+
data = [1, 2, 3, 4, 500, 525, 602, 630, 1000, 1100, 1065, 1150]
|
31
|
+
apc = AffinityPropagation::Calculator.new(data) do |datum, exemplar|
|
32
|
+
# A recommended similarity function from the original paper:
|
33
|
+
# negative square of the distance between two points
|
34
|
+
# However you can choose anything here and even make it asymmetrical if your data needs it
|
35
|
+
-((datum - exemplar) ** 2)
|
36
|
+
end
|
37
|
+
apc.run(stable_iterations: 50)
|
38
|
+
apc.clusters
|
39
|
+
```
|
40
|
+
This will result in the following (beautified) output:
|
41
|
+
```ruby
|
42
|
+
[
|
43
|
+
{
|
44
|
+
:exemplar=>2,
|
45
|
+
:members=>[1, 2, 4]
|
46
|
+
},
|
47
|
+
{
|
48
|
+
:exemplar=>3,
|
49
|
+
:members=>[3]
|
50
|
+
},
|
51
|
+
{
|
52
|
+
:exemplar=>602,
|
53
|
+
:members=>[500, 525, 602, 630]
|
54
|
+
},
|
55
|
+
{
|
56
|
+
:exemplar=>1065,
|
57
|
+
:members=>[1000, 1100, 1065, 1150]
|
58
|
+
}
|
59
|
+
]
|
60
|
+
```
|
61
|
+
A more real-world example would be something like the following:
|
62
|
+
```ruby
|
63
|
+
require 'active_support/all'
|
64
|
+
data = [
|
65
|
+
{ value: 'A', timestamp: 1.year.ago },
|
66
|
+
{ value: 'B', timestamp: 1.year.ago },
|
67
|
+
{ value: 'C', timestamp: 1.year.ago },
|
68
|
+
{ value: 'D', timestamp: 1.year.ago + 1.hour },
|
69
|
+
{ value: 'E', timestamp: 1.year.ago + 1.day },
|
70
|
+
|
71
|
+
{ value: 'a', timestamp: 1.month.ago },
|
72
|
+
{ value: 'b', timestamp: 1.month.ago + 1.hour },
|
73
|
+
{ value: 'c', timestamp: 1.month.ago + 1.day },
|
74
|
+
|
75
|
+
{ value: 'd', timestamp: 2.weeks.ago },
|
76
|
+
{ value: 'e', timestamp: 2.weeks.ago - 3.hours },
|
77
|
+
{ value: 'f', timestamp: 2.weeks.ago - 10.hours },
|
78
|
+
{ value: 'g', timestamp: 2.weeks.ago + 5.hours },
|
79
|
+
|
80
|
+
{ value: 'h', timestamp: 24.hours.ago },
|
81
|
+
{ value: 'i', timestamp: 40.hours.ago },
|
82
|
+
{ value: 'j', timestamp: 12.hours.ago },
|
83
|
+
{ value: 'k', timestamp: 12.minutes.ago }
|
84
|
+
]
|
85
|
+
apc = AffinityPropagation::Calculator.new(data, lambda: 0.9) do |datum, exemplar|
|
86
|
+
difference = datum[:timestamp] - exemplar[:timestamp]
|
87
|
+
|
88
|
+
-(difference ** 2)
|
89
|
+
end
|
90
|
+
apc.run(stable_iterations: 25)
|
91
|
+
apc.clusters.map { |center| center[:members].map { |item| item[:value] } }
|
92
|
+
```
|
93
|
+
## Tweaking the Result
|
94
|
+
There are at least two ways to tweak the algorithm to generate clusters closer to any expectations you might have for
|
95
|
+
your data:
|
96
|
+
|
97
|
+
1. Changing the dampening factor, `lambda`, which can be provided to the constructor for `AffinityPropagation::Calculator`.
|
98
|
+
This helps prevent numerical oscillations while updating the responsibilities and similarities matrices. This defaults
|
99
|
+
0.75 but can be changed by specifying the lambda name argument as in the second example above.
|
100
|
+
1. Assigning custom values to the self-similarity nodes in the similarity matrix generated when you instantiate the
|
101
|
+
`AffinityPropagation::Calculator` object. These self-similarity values are currently assigned a common value to
|
102
|
+
not suggest any one exemplar over another. This common value is the median of all similarities found in the entire
|
103
|
+
similarity matrix. This is not currently tweakable from your code but there's a plan to allow this to be tweaked.
|
104
|
+
|
105
|
+
## Development
|
106
|
+
|
107
|
+
After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment.
|
108
|
+
|
109
|
+
To install this gem onto your local machine, run `bundle exec rake install`. To release a new version, update the version number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git commits and tags, and push the `.gem` file to [rubygems.org](https://rubygems.org).
|
110
|
+
|
111
|
+
## Contributing
|
112
|
+
|
113
|
+
This is my initial take on this problem and I have still to add a few usability, configuration and performance tweaks
|
114
|
+
but wanted to get this out there so people can use it. If you find it useful and are able to contribute to remove some
|
115
|
+
of the clearly vast set of issues that this codebase has, please send me a pull request and let's talk!
|
116
|
+
|
117
|
+
Bug reports and pull requests are welcome on GitHub at https://github.com/sjaveed/affinity_propagation.
|
118
|
+
|
119
|
+
## License
|
120
|
+
|
121
|
+
The gem is available as open source under the terms of the [MIT License](http://opensource.org/licenses/MIT).
|
data/Rakefile
ADDED
@@ -0,0 +1,32 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
lib = File.expand_path("../lib", __FILE__)
|
3
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
4
|
+
require "affinity_propagation/version"
|
5
|
+
|
6
|
+
Gem::Specification.new do |spec|
|
7
|
+
spec.name = "affinity_propagation"
|
8
|
+
spec.version = AffinityPropagation::VERSION
|
9
|
+
spec.authors = ["Shahbaz Javeed"]
|
10
|
+
spec.email = ["sjaveed@gmail.com"]
|
11
|
+
|
12
|
+
spec.summary = %q{An implementation of the affinity propagation clustering algorithm by Frey and Dueck}
|
13
|
+
spec.description = %q{
|
14
|
+
Affinity Propagation is a clustering algorithm that does not require pre-specifying the number of clusters like
|
15
|
+
k-means and other similar algorithms do. This is a ruby implementation of the original version defined by Frey and
|
16
|
+
Dueck in http://www.psi.toronto.edu/affinitypropagation/FreyDueckScience07.pdf
|
17
|
+
}
|
18
|
+
spec.homepage = "https://github.com/sjaveed/affinity_propagation"
|
19
|
+
spec.license = "MIT"
|
20
|
+
|
21
|
+
spec.files = `git ls-files -z`.split("\x0").reject do |f|
|
22
|
+
f.match(%r{^(test|spec|features)/})
|
23
|
+
end
|
24
|
+
spec.bindir = "exe"
|
25
|
+
spec.executables = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
|
26
|
+
spec.require_paths = ["lib"]
|
27
|
+
|
28
|
+
spec.add_development_dependency "bundler", "~> 1.15"
|
29
|
+
spec.add_development_dependency "rake", "~> 10.0"
|
30
|
+
spec.add_development_dependency "rspec", "~> 3.0"
|
31
|
+
spec.add_development_dependency "activesupport"
|
32
|
+
end
|
data/bin/console
ADDED
@@ -0,0 +1,14 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
require "bundler/setup"
|
4
|
+
require "affinity_propagation"
|
5
|
+
|
6
|
+
# You can add fixtures and/or initialization code here to make experimenting
|
7
|
+
# with your gem easier. You can also use a different console, if you like.
|
8
|
+
|
9
|
+
# (If you use this, don't forget to add pry to your Gemfile!)
|
10
|
+
# require "pry"
|
11
|
+
# Pry.start
|
12
|
+
|
13
|
+
require "irb"
|
14
|
+
IRB.start(__FILE__)
|
data/bin/setup
ADDED
@@ -0,0 +1,180 @@
|
|
1
|
+
require 'matrix'
|
2
|
+
|
3
|
+
module AffinityPropagation
|
4
|
+
refine Array do
|
5
|
+
def median
|
6
|
+
relevant_elements = if size % 2 == 0
|
7
|
+
# Even number of items in this list => let's get the middle two and return their mean
|
8
|
+
slice(size / 2, 2)
|
9
|
+
else
|
10
|
+
slice(size / 2, 1)
|
11
|
+
end
|
12
|
+
|
13
|
+
relevant_elements.sum / relevant_elements.size
|
14
|
+
end
|
15
|
+
end
|
16
|
+
|
17
|
+
refine Matrix do
|
18
|
+
def to_matlab
|
19
|
+
output = row_vectors.map(&:to_a).map { |row| row.join(', ') }.join('; ')
|
20
|
+
|
21
|
+
"[#{output}]"
|
22
|
+
end
|
23
|
+
end
|
24
|
+
|
25
|
+
class Calculator
|
26
|
+
using AffinityPropagation
|
27
|
+
|
28
|
+
LAMBDA = 0.75
|
29
|
+
|
30
|
+
attr_accessor :raw_clusters
|
31
|
+
|
32
|
+
def initialize(data, lambda: LAMBDA, &block)
|
33
|
+
@data = data
|
34
|
+
@lambda = lambda
|
35
|
+
|
36
|
+
raise 'no block provide to calculate similarities within data!' unless block_given?
|
37
|
+
|
38
|
+
@similarities = similarity_matrix(&block)
|
39
|
+
reset
|
40
|
+
end
|
41
|
+
|
42
|
+
def reset
|
43
|
+
@raw_clusters = {}
|
44
|
+
@stable_cluster_iterations = 0
|
45
|
+
@total_iterations = 0
|
46
|
+
|
47
|
+
@availabilities = Matrix.zero(@data.size, @data.size)
|
48
|
+
@responsibilities = Matrix.zero(@data.size, @data.size)
|
49
|
+
end
|
50
|
+
|
51
|
+
def iterate
|
52
|
+
@availabilities = availability_matrix
|
53
|
+
@responsibilities = responsibility_matrix
|
54
|
+
update_clusters
|
55
|
+
|
56
|
+
@total_iterations += 1
|
57
|
+
end
|
58
|
+
|
59
|
+
def clusters
|
60
|
+
clusters = []
|
61
|
+
|
62
|
+
raw_clusters.each do |exemplar_id, data_ids|
|
63
|
+
clusters << {
|
64
|
+
exemplar: @data[exemplar_id],
|
65
|
+
members: data_ids.map { |datum_id| @data[datum_id] }
|
66
|
+
}
|
67
|
+
end
|
68
|
+
|
69
|
+
clusters
|
70
|
+
end
|
71
|
+
|
72
|
+
def run(iterations: 100, stable_iterations: 10)
|
73
|
+
iterate while @total_iterations < iterations && @stable_cluster_iterations < stable_iterations
|
74
|
+
end
|
75
|
+
|
76
|
+
private
|
77
|
+
|
78
|
+
def similarity_matrix(&block)
|
79
|
+
similarity_array = []
|
80
|
+
similarities = Matrix.build(@data.size, @data.size) do |row_idx, col_idx|
|
81
|
+
exemplar = @data[row_idx]
|
82
|
+
datum = @data[col_idx]
|
83
|
+
|
84
|
+
similarity = block.call(datum, exemplar)
|
85
|
+
similarity_array << similarity
|
86
|
+
similarity
|
87
|
+
end
|
88
|
+
|
89
|
+
median_similarity = similarity_array.median
|
90
|
+
|
91
|
+
# Matrices can't be modified once created so we have to use this hack
|
92
|
+
(0...@data.size).each { |idx| similarities.send(:[]=, idx, idx, median_similarity) }
|
93
|
+
|
94
|
+
similarities
|
95
|
+
end
|
96
|
+
|
97
|
+
def dampen(new_value, existing_value)
|
98
|
+
(1 - @lambda) * new_value + @lambda * existing_value
|
99
|
+
end
|
100
|
+
|
101
|
+
def responsibility_matrix
|
102
|
+
Matrix.build(@similarities.row_count, @similarities.column_count) do |row_idx, col_idx|
|
103
|
+
exemplar_idx = row_idx
|
104
|
+
datum_idx = col_idx
|
105
|
+
|
106
|
+
availability_column = @availabilities.column(col_idx).to_a
|
107
|
+
availability_column.slice!(exemplar_idx)
|
108
|
+
similarity_column = @similarities.column(col_idx).to_a
|
109
|
+
similarity_column.slice!(exemplar_idx)
|
110
|
+
|
111
|
+
availability_plus_similarity = []
|
112
|
+
availability_column.zip(similarity_column) { |data| availability_plus_similarity << data.sum }
|
113
|
+
|
114
|
+
dampen(@similarities[row_idx, col_idx] - availability_plus_similarity.max, @responsibilities[exemplar_idx, datum_idx])
|
115
|
+
end
|
116
|
+
end
|
117
|
+
|
118
|
+
def availability_matrix
|
119
|
+
Matrix.build(@responsibilities.row_count, @responsibilities.column_count) do |row_idx, col_idx|
|
120
|
+
exemplar_idx = row_idx
|
121
|
+
datum_idx = col_idx
|
122
|
+
responsibility_column = @responsibilities.row(exemplar_idx).to_a
|
123
|
+
|
124
|
+
if exemplar_idx == datum_idx
|
125
|
+
# self-availability
|
126
|
+
responsibility_column.slice!(exemplar_idx)
|
127
|
+
|
128
|
+
dampen(responsibility_column.inject(0) { |sum, item| sum += [0, item].max },@availabilities[exemplar_idx, datum_idx])
|
129
|
+
else
|
130
|
+
self_responsibility = @responsibilities[exemplar_idx, exemplar_idx]
|
131
|
+
if datum_idx > exemplar_idx
|
132
|
+
# Slice out the datum index first since in this case it won't affect the exemplar index
|
133
|
+
responsibility_column.slice!(datum_idx)
|
134
|
+
responsibility_column.slice!(exemplar_idx)
|
135
|
+
else
|
136
|
+
responsibility_column.slice!(exemplar_idx)
|
137
|
+
responsibility_column.slice!(datum_idx)
|
138
|
+
end
|
139
|
+
|
140
|
+
responsibility_column_sum = responsibility_column.inject(0) { |sum, item| sum += [0, item].max }
|
141
|
+
|
142
|
+
dampen([0, self_responsibility + responsibility_column_sum].min, @availabilities[exemplar_idx, datum_idx])
|
143
|
+
end
|
144
|
+
end
|
145
|
+
end
|
146
|
+
|
147
|
+
def identify_raw_clusters
|
148
|
+
clusters = {}
|
149
|
+
|
150
|
+
@data.each_with_index do |item, datum_idx|
|
151
|
+
availability_column = @availabilities.column(datum_idx).to_a
|
152
|
+
responsibility_column = @responsibilities.column(datum_idx).to_a
|
153
|
+
|
154
|
+
availability_and_responsibility = @data.size.times.map do |exemplar_idx|
|
155
|
+
availability_column[exemplar_idx] + responsibility_column[exemplar_idx]
|
156
|
+
end
|
157
|
+
exemplar_idx = availability_and_responsibility.index(availability_and_responsibility.max)
|
158
|
+
|
159
|
+
if clusters.key?(exemplar_idx)
|
160
|
+
clusters[exemplar_idx] << datum_idx
|
161
|
+
else
|
162
|
+
clusters[exemplar_idx] = [datum_idx]
|
163
|
+
end
|
164
|
+
end
|
165
|
+
|
166
|
+
clusters
|
167
|
+
end
|
168
|
+
|
169
|
+
def update_clusters
|
170
|
+
new_clusters = identify_raw_clusters
|
171
|
+
|
172
|
+
if new_clusters == @raw_clusters
|
173
|
+
@stable_cluster_iterations += 1
|
174
|
+
else
|
175
|
+
@raw_clusters = new_clusters
|
176
|
+
@stable_cluster_iterations = 0
|
177
|
+
end
|
178
|
+
end
|
179
|
+
end
|
180
|
+
end
|
metadata
ADDED
@@ -0,0 +1,121 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: affinity_propagation
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.1.0
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Shahbaz Javeed
|
8
|
+
autorequire:
|
9
|
+
bindir: exe
|
10
|
+
cert_chain: []
|
11
|
+
date: 2017-09-10 00:00:00.000000000 Z
|
12
|
+
dependencies:
|
13
|
+
- !ruby/object:Gem::Dependency
|
14
|
+
name: bundler
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
16
|
+
requirements:
|
17
|
+
- - "~>"
|
18
|
+
- !ruby/object:Gem::Version
|
19
|
+
version: '1.15'
|
20
|
+
type: :development
|
21
|
+
prerelease: false
|
22
|
+
version_requirements: !ruby/object:Gem::Requirement
|
23
|
+
requirements:
|
24
|
+
- - "~>"
|
25
|
+
- !ruby/object:Gem::Version
|
26
|
+
version: '1.15'
|
27
|
+
- !ruby/object:Gem::Dependency
|
28
|
+
name: rake
|
29
|
+
requirement: !ruby/object:Gem::Requirement
|
30
|
+
requirements:
|
31
|
+
- - "~>"
|
32
|
+
- !ruby/object:Gem::Version
|
33
|
+
version: '10.0'
|
34
|
+
type: :development
|
35
|
+
prerelease: false
|
36
|
+
version_requirements: !ruby/object:Gem::Requirement
|
37
|
+
requirements:
|
38
|
+
- - "~>"
|
39
|
+
- !ruby/object:Gem::Version
|
40
|
+
version: '10.0'
|
41
|
+
- !ruby/object:Gem::Dependency
|
42
|
+
name: rspec
|
43
|
+
requirement: !ruby/object:Gem::Requirement
|
44
|
+
requirements:
|
45
|
+
- - "~>"
|
46
|
+
- !ruby/object:Gem::Version
|
47
|
+
version: '3.0'
|
48
|
+
type: :development
|
49
|
+
prerelease: false
|
50
|
+
version_requirements: !ruby/object:Gem::Requirement
|
51
|
+
requirements:
|
52
|
+
- - "~>"
|
53
|
+
- !ruby/object:Gem::Version
|
54
|
+
version: '3.0'
|
55
|
+
- !ruby/object:Gem::Dependency
|
56
|
+
name: activesupport
|
57
|
+
requirement: !ruby/object:Gem::Requirement
|
58
|
+
requirements:
|
59
|
+
- - ">="
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: '0'
|
62
|
+
type: :development
|
63
|
+
prerelease: false
|
64
|
+
version_requirements: !ruby/object:Gem::Requirement
|
65
|
+
requirements:
|
66
|
+
- - ">="
|
67
|
+
- !ruby/object:Gem::Version
|
68
|
+
version: '0'
|
69
|
+
description: "\n Affinity Propagation is a clustering algorithm that does not require
|
70
|
+
pre-specifying the number of clusters like\n k-means and other similar algorithms
|
71
|
+
do. This is a ruby implementation of the original version defined by Frey and\n
|
72
|
+
\ Dueck in http://www.psi.toronto.edu/affinitypropagation/FreyDueckScience07.pdf\n
|
73
|
+
\ "
|
74
|
+
email:
|
75
|
+
- sjaveed@gmail.com
|
76
|
+
executables: []
|
77
|
+
extensions: []
|
78
|
+
extra_rdoc_files: []
|
79
|
+
files:
|
80
|
+
- ".gitignore"
|
81
|
+
- ".rspec"
|
82
|
+
- ".ruby-gemset"
|
83
|
+
- ".ruby-version"
|
84
|
+
- ".travis.yml"
|
85
|
+
- Gemfile
|
86
|
+
- Gemfile.lock
|
87
|
+
- LICENSE.txt
|
88
|
+
- README.md
|
89
|
+
- Rakefile
|
90
|
+
- affinity_propagation.gemspec
|
91
|
+
- bin/console
|
92
|
+
- bin/setup
|
93
|
+
- lib/affinity_propagation.rb
|
94
|
+
- lib/affinity_propagation/calculator.rb
|
95
|
+
- lib/affinity_propagation/version.rb
|
96
|
+
homepage: https://github.com/sjaveed/affinity_propagation
|
97
|
+
licenses:
|
98
|
+
- MIT
|
99
|
+
metadata: {}
|
100
|
+
post_install_message:
|
101
|
+
rdoc_options: []
|
102
|
+
require_paths:
|
103
|
+
- lib
|
104
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
105
|
+
requirements:
|
106
|
+
- - ">="
|
107
|
+
- !ruby/object:Gem::Version
|
108
|
+
version: '0'
|
109
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
110
|
+
requirements:
|
111
|
+
- - ">="
|
112
|
+
- !ruby/object:Gem::Version
|
113
|
+
version: '0'
|
114
|
+
requirements: []
|
115
|
+
rubyforge_project:
|
116
|
+
rubygems_version: 2.6.11
|
117
|
+
signing_key:
|
118
|
+
specification_version: 4
|
119
|
+
summary: An implementation of the affinity propagation clustering algorithm by Frey
|
120
|
+
and Dueck
|
121
|
+
test_files: []
|