clusta 0.0.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/LICENSE +20 -0
- data/README.rdoc +0 -0
- data/VERSION +1 -0
- data/bin/clusta +35 -0
- data/lib/clusta.rb +24 -0
- data/lib/clusta/geometry.rb +25 -0
- data/lib/clusta/geometry/arrow.rb +16 -0
- data/lib/clusta/geometry/assortativity.rb +31 -0
- data/lib/clusta/geometry/degree.rb +28 -0
- data/lib/clusta/geometry/directed/arrow.rb +12 -0
- data/lib/clusta/geometry/directed/degree.rb +29 -0
- data/lib/clusta/geometry/directed/edge.rb +26 -0
- data/lib/clusta/geometry/directed/edge_degree_pair.rb +24 -0
- data/lib/clusta/geometry/directed/vertex_arrows.rb +25 -0
- data/lib/clusta/geometry/edge.rb +58 -0
- data/lib/clusta/geometry/edge_degree_pair.rb +21 -0
- data/lib/clusta/geometry/element.rb +131 -0
- data/lib/clusta/geometry/vertex.rb +11 -0
- data/lib/clusta/geometry/vertex_arrows.rb +45 -0
- data/lib/clusta/runner.rb +15 -0
- data/lib/clusta/transforms.rb +50 -0
- data/lib/clusta/transforms/edge_degree_pairs_to_assortativities.rb +42 -0
- data/lib/clusta/transforms/edges_to_degrees.rb +42 -0
- data/lib/clusta/transforms/edges_to_vertex_arrows.rb +50 -0
- data/lib/clusta/transforms/import.rb +17 -0
- data/lib/clusta/transforms/vertex_arrows_to_edge_degree_pairs.rb +63 -0
- data/spec/clusta/geometry/element_spec.rb +191 -0
- data/spec/clusta/transforms/edges_to_degrees_spec.rb +22 -0
- data/spec/clusta/transforms/edges_to_vertex_arrows_spec.rb +21 -0
- data/spec/data/README.rdoc +54 -0
- data/spec/data/degrees/directed.tsv +9 -0
- data/spec/data/degrees/undirected.tsv +9 -0
- data/spec/data/edges/directed.unweighted.tsv +10 -0
- data/spec/data/edges/directed.weighted.tsv +10 -0
- data/spec/data/edges/undirected.unweighted.tsv +9 -0
- data/spec/data/edges/undirected.weighted.tsv +9 -0
- data/spec/data/vertex_arrows/directed.unweighted.tsv +7 -0
- data/spec/data/vertex_arrows/directed.weighted.tsv +7 -0
- data/spec/data/vertex_arrows/undirected.unweighted.tsv +9 -0
- data/spec/data/vertex_arrows/undirected.weighted.tsv +9 -0
- data/spec/spec_helper.rb +21 -0
- data/spec/support/transforms_spec_helper.rb +120 -0
- metadata +123 -0
data/LICENSE
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
Copyright (c) 2012 Dhruv Bansal
|
2
|
+
|
3
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
4
|
+
a copy of this software and associated documentation files (the
|
5
|
+
"Software"), to deal in the Software without restriction, including
|
6
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
7
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
8
|
+
permit persons to whom the Software is furnished to do so, subject to
|
9
|
+
the following conditions:
|
10
|
+
|
11
|
+
The above copyright notice and this permission notice shall be
|
12
|
+
included in all copies or substantial portions of the Software.
|
13
|
+
|
14
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
15
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
16
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
17
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
18
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
19
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
20
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.rdoc
ADDED
File without changes
|
data/VERSION
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
0.0.1
|
data/bin/clusta
ADDED
@@ -0,0 +1,35 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
$: << File.expand_path('../lib', File.dirname(__FILE__)) unless $:.include?(File.expand_path('../lib', File.dirname(__FILE__)))
|
4
|
+
|
5
|
+
require 'clusta'
|
6
|
+
|
7
|
+
def usage
|
8
|
+
"usage: #{File.basename(__FILE__)} --transform=TRANSFORM_NAME [ARGS ...]"
|
9
|
+
end
|
10
|
+
|
11
|
+
def extract_transform_arg
|
12
|
+
transform_arg = ARGV.find_all { |arg| arg =~ Clusta::Transforms::ARG_REGEXP }.first
|
13
|
+
if transform_arg.nil?
|
14
|
+
$stderr.puts(usage)
|
15
|
+
exit(1)
|
16
|
+
end
|
17
|
+
# ARGV.delete_if { |arg| arg =~ Clusta::Transforms::ARG_REGEXP }
|
18
|
+
transform_arg
|
19
|
+
end
|
20
|
+
|
21
|
+
def add_default_run_arg
|
22
|
+
ARGV.unshift('--run=local') unless ARGV.detect { |arg| arg =~ /--run/ }
|
23
|
+
end
|
24
|
+
|
25
|
+
if $0 == __FILE__
|
26
|
+
begin
|
27
|
+
add_default_run_arg
|
28
|
+
transform = Clusta::Transforms.from_arg(extract_transform_arg)
|
29
|
+
script = Clusta::Transforms.script_for(transform)
|
30
|
+
script.run
|
31
|
+
rescue Clusta::Error => e
|
32
|
+
$stderr.puts e.message
|
33
|
+
exit(1)
|
34
|
+
end
|
35
|
+
end
|
data/lib/clusta.rb
ADDED
@@ -0,0 +1,24 @@
|
|
1
|
+
require 'wukong'
|
2
|
+
|
3
|
+
module Clusta
|
4
|
+
|
5
|
+
def self.underscore string
|
6
|
+
string.gsub(/([A-Z])/, '_\1').downcase[1..-1]
|
7
|
+
end
|
8
|
+
|
9
|
+
def self.classify string
|
10
|
+
string.split('_').map(&:capitalize).join
|
11
|
+
end
|
12
|
+
|
13
|
+
def self.require_name path
|
14
|
+
File.basename(path).gsub(/\.rb$/, '')
|
15
|
+
end
|
16
|
+
|
17
|
+
Error = Class.new(StandardError)
|
18
|
+
DirectednessMismatchError = Class.new(Error)
|
19
|
+
AmbiguousArgumentsError = Class.new(Error)
|
20
|
+
|
21
|
+
autoload :Geometry, 'clusta/geometry'
|
22
|
+
autoload :Transforms, 'clusta/transforms'
|
23
|
+
|
24
|
+
end
|
@@ -0,0 +1,25 @@
|
|
1
|
+
module Clusta
|
2
|
+
|
3
|
+
module Geometry
|
4
|
+
|
5
|
+
autoload :Element, 'clusta/geometry/element'
|
6
|
+
|
7
|
+
ELEMENTS = []
|
8
|
+
|
9
|
+
def self.register_geometry name, path, geometries=nil
|
10
|
+
autoload name, path
|
11
|
+
self::ELEMENTS << name
|
12
|
+
end
|
13
|
+
|
14
|
+
Dir[File.join(File.dirname(__FILE__), "geometry/*.rb")].each do |path|
|
15
|
+
require_name = Clusta.require_name(path)
|
16
|
+
register_geometry Clusta.classify(require_name), "clusta/geometry/#{require_name}"
|
17
|
+
end
|
18
|
+
|
19
|
+
Dir[File.join(File.dirname(__FILE__), "geometry/directed/*.rb")].each do |path|
|
20
|
+
require_name = Clusta.require_name(path)
|
21
|
+
register_geometry ("Directed" + Clusta.classify(require_name)), "clusta/geometry/directed/#{require_name}"
|
22
|
+
end
|
23
|
+
|
24
|
+
end
|
25
|
+
end
|
@@ -0,0 +1,31 @@
|
|
1
|
+
module Clusta
|
2
|
+
|
3
|
+
module Geometry
|
4
|
+
|
5
|
+
class Assortativity < Element
|
6
|
+
|
7
|
+
field :source_degree_value, :type => :int
|
8
|
+
field :target_degree_value, :type => :int
|
9
|
+
field :count, :type => :int
|
10
|
+
|
11
|
+
def directed?
|
12
|
+
false
|
13
|
+
end
|
14
|
+
|
15
|
+
def key
|
16
|
+
[source_degree_value, target_degree_value]
|
17
|
+
end
|
18
|
+
|
19
|
+
def zero
|
20
|
+
self.class.new(*(key + [0]))
|
21
|
+
end
|
22
|
+
|
23
|
+
def +(other)
|
24
|
+
raise DirectednessMismatchError.new if other.directed?
|
25
|
+
self.class.new(*(key + [count + other.count]))
|
26
|
+
end
|
27
|
+
|
28
|
+
end
|
29
|
+
|
30
|
+
end
|
31
|
+
end
|
@@ -0,0 +1,28 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class Degree < Element
|
5
|
+
|
6
|
+
field :vertex_label
|
7
|
+
field :degree, :type => :int
|
8
|
+
|
9
|
+
def directed?
|
10
|
+
false
|
11
|
+
end
|
12
|
+
|
13
|
+
def zero
|
14
|
+
self.class.new(vertex_label, 0)
|
15
|
+
end
|
16
|
+
|
17
|
+
def one deg=1
|
18
|
+
self.class.new(vertex_label, deg)
|
19
|
+
end
|
20
|
+
|
21
|
+
def +(other)
|
22
|
+
raise DirectednessMismatchError.new if other.directed?
|
23
|
+
self.class.new(vertex_label, self.degree + other.degree)
|
24
|
+
end
|
25
|
+
end
|
26
|
+
|
27
|
+
end
|
28
|
+
end
|
@@ -0,0 +1,29 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class DirectedDegree < Element
|
5
|
+
|
6
|
+
field :vertex_label
|
7
|
+
field :in_degree, :type => :int
|
8
|
+
field :out_degree, :type => :int
|
9
|
+
|
10
|
+
def directed?
|
11
|
+
true
|
12
|
+
end
|
13
|
+
|
14
|
+
def zero
|
15
|
+
self.class.new(vertex_label, 0, 0)
|
16
|
+
end
|
17
|
+
|
18
|
+
def one inbound=1, outbound=0
|
19
|
+
self.class.new(vertex_label, inbound, outbound)
|
20
|
+
end
|
21
|
+
|
22
|
+
def +(other)
|
23
|
+
raise DirectednessMismatchError.new unless other.directed?
|
24
|
+
self.class.new(vertex_label, self.in_degree + other.in_degree, self.out_degree + other.out_degree)
|
25
|
+
end
|
26
|
+
end
|
27
|
+
|
28
|
+
end
|
29
|
+
end
|
@@ -0,0 +1,26 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class DirectedEdge < Edge
|
5
|
+
|
6
|
+
def directed?
|
7
|
+
true
|
8
|
+
end
|
9
|
+
|
10
|
+
def source_degree
|
11
|
+
DirectedDegree.new(source_label, 0, 1)
|
12
|
+
end
|
13
|
+
|
14
|
+
def target_degree
|
15
|
+
DirectedDegree.new(target_label, 1, 0)
|
16
|
+
end
|
17
|
+
|
18
|
+
def arrow
|
19
|
+
DirectedArrow.new(target_label, weight)
|
20
|
+
end
|
21
|
+
|
22
|
+
|
23
|
+
end
|
24
|
+
|
25
|
+
end
|
26
|
+
end
|
@@ -0,0 +1,24 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class DirectedEdgeDegreePair < Element
|
5
|
+
field :source_label
|
6
|
+
field :target_label
|
7
|
+
field :source_in_degree_value, :type => :int
|
8
|
+
field :source_out_degree_value, :type => :int
|
9
|
+
field :target_in_degree_value, :type => :int
|
10
|
+
field :target_out_degree_value, :type => :int
|
11
|
+
field :weight, :optional => true
|
12
|
+
|
13
|
+
def directed?
|
14
|
+
true
|
15
|
+
end
|
16
|
+
|
17
|
+
def assortativity
|
18
|
+
Assortativity.new(source_in_degree_value, target_out_degree_value, 1)
|
19
|
+
end
|
20
|
+
|
21
|
+
end
|
22
|
+
|
23
|
+
end
|
24
|
+
end
|
@@ -0,0 +1,25 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class DirectedVertexArrows < VertexArrows
|
5
|
+
|
6
|
+
def directed?
|
7
|
+
true
|
8
|
+
end
|
9
|
+
|
10
|
+
def source_degrees
|
11
|
+
[0, arrows.size]
|
12
|
+
end
|
13
|
+
|
14
|
+
def target_degrees
|
15
|
+
[1, 0]
|
16
|
+
end
|
17
|
+
|
18
|
+
def edge_degree_pair source_label, target_label, *args
|
19
|
+
DirectedEdgeDegreePair.new(source_label, target_label, *args)
|
20
|
+
end
|
21
|
+
|
22
|
+
end
|
23
|
+
|
24
|
+
end
|
25
|
+
end
|
@@ -0,0 +1,58 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class Edge < Element
|
5
|
+
|
6
|
+
field :source_label
|
7
|
+
field :target_label
|
8
|
+
field :weight, :optional => true
|
9
|
+
|
10
|
+
def weighted?
|
11
|
+
self.weight
|
12
|
+
end
|
13
|
+
|
14
|
+
def directed?
|
15
|
+
false
|
16
|
+
end
|
17
|
+
|
18
|
+
def joins? label
|
19
|
+
source_label == label || target_label == label
|
20
|
+
end
|
21
|
+
|
22
|
+
def labels_string
|
23
|
+
[source_label, target_label].map(&:to_s).join(' -> ')
|
24
|
+
end
|
25
|
+
|
26
|
+
def source_degree
|
27
|
+
Degree.new(source_label, 1)
|
28
|
+
end
|
29
|
+
|
30
|
+
def target_degree
|
31
|
+
Degree.new(target_label, 1)
|
32
|
+
end
|
33
|
+
|
34
|
+
def degrees
|
35
|
+
[source_degree, target_degree]
|
36
|
+
end
|
37
|
+
|
38
|
+
def degree_of label
|
39
|
+
case label
|
40
|
+
when source_label then source_degree
|
41
|
+
when target_label then target_degree
|
42
|
+
else
|
43
|
+
raise Error.new("This edge (#{labels_string}) does not contain vertex #{label}")
|
44
|
+
end
|
45
|
+
end
|
46
|
+
|
47
|
+
def reversed
|
48
|
+
self.class.new(target_label, source_label, weight)
|
49
|
+
end
|
50
|
+
|
51
|
+
def arrow
|
52
|
+
Arrow.new(target_label, weight)
|
53
|
+
end
|
54
|
+
|
55
|
+
end
|
56
|
+
|
57
|
+
end
|
58
|
+
end
|
@@ -0,0 +1,21 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class EdgeDegreePair < Element
|
5
|
+
field :source_label
|
6
|
+
field :target_label
|
7
|
+
field :source_degree_value, :type => :int
|
8
|
+
field :target_degree_value, :type => :int
|
9
|
+
|
10
|
+
def directed?
|
11
|
+
false
|
12
|
+
end
|
13
|
+
|
14
|
+
def assortativity
|
15
|
+
Assortativity.new(source_degree_value, target_degree_value, 1)
|
16
|
+
end
|
17
|
+
|
18
|
+
end
|
19
|
+
|
20
|
+
end
|
21
|
+
end
|
@@ -0,0 +1,131 @@
|
|
1
|
+
module Clusta
|
2
|
+
module Geometry
|
3
|
+
|
4
|
+
class Element
|
5
|
+
|
6
|
+
attr_accessor :input_fields
|
7
|
+
|
8
|
+
@fields = []
|
9
|
+
class << self ; attr_reader :fields ; end
|
10
|
+
|
11
|
+
def self.inherited(subclass)
|
12
|
+
subclass.instance_variable_set("@fields", @fields.dup)
|
13
|
+
super
|
14
|
+
end
|
15
|
+
|
16
|
+
def self.field_names
|
17
|
+
@fields.map { |field| field[:name].to_s }
|
18
|
+
end
|
19
|
+
|
20
|
+
def self.has_optional_field?
|
21
|
+
@fields.any? { |field| field[:optional] }
|
22
|
+
end
|
23
|
+
|
24
|
+
def self.optional_field
|
25
|
+
@fields.detect { |field| field[:optional] }
|
26
|
+
end
|
27
|
+
|
28
|
+
def self.from_string string
|
29
|
+
return string unless string.is_a?(String)
|
30
|
+
args = string.split(';')
|
31
|
+
klass_name = args.shift
|
32
|
+
raise ArgumentError.new("Elements instantiated from a string must match the format 'klass;[field1;[field2;]...]'") unless klass_name
|
33
|
+
Wukong.class_from_resource(klass_name).new(*args)
|
34
|
+
end
|
35
|
+
|
36
|
+
def self.field name, options={}
|
37
|
+
raise AmbiguousArgumentsError.new("Cannot define a second optional field #{name} because field #{optional_field[:name]} is already optional.") if has_optional_field?
|
38
|
+
attr_reader name
|
39
|
+
case options[:type]
|
40
|
+
when :int
|
41
|
+
define_method "#{name}=" do |val|
|
42
|
+
instance_variable_set("@#{name}", val.to_i)
|
43
|
+
end
|
44
|
+
when :float
|
45
|
+
define_method "#{name}=" do |val|
|
46
|
+
instance_variable_set("@#{name}", val.to_f)
|
47
|
+
end
|
48
|
+
when :geometry
|
49
|
+
define_method "#{name}=" do |val|
|
50
|
+
instance_variable_set("@#{name}", self.class.from_string(val))
|
51
|
+
end
|
52
|
+
else
|
53
|
+
define_method "#{name}=" do |val|
|
54
|
+
instance_variable_set("@#{name}", val)
|
55
|
+
end
|
56
|
+
end
|
57
|
+
@fields << options.merge(:name => name)
|
58
|
+
end
|
59
|
+
|
60
|
+
def fields
|
61
|
+
self.class.fields
|
62
|
+
end
|
63
|
+
|
64
|
+
def self.input_fields name
|
65
|
+
alias_method name, :input_fields
|
66
|
+
end
|
67
|
+
|
68
|
+
def self.stream_name
|
69
|
+
if defined?(Settings) && Settings[:full_class_names]
|
70
|
+
to_s
|
71
|
+
else
|
72
|
+
to_s.split("::").last
|
73
|
+
end
|
74
|
+
end
|
75
|
+
|
76
|
+
def stream_name
|
77
|
+
self.class.stream_name
|
78
|
+
end
|
79
|
+
|
80
|
+
def initialize *args
|
81
|
+
self.class.fields.each_with_index do |field, index|
|
82
|
+
suffix = case index.to_s
|
83
|
+
when /1$/ then 'st'
|
84
|
+
when /2$/ then 'nd'
|
85
|
+
when /3$/ then 'rd'
|
86
|
+
else 'th'
|
87
|
+
end
|
88
|
+
case
|
89
|
+
when field[:optional]
|
90
|
+
self.send("#{field[:name]}=", args[index]) if args[index]
|
91
|
+
when args[index].nil?
|
92
|
+
raise ArgumentError.new("A #{self.class} requires a non-nil value for #{field[:name]} as its #{index}#{suffix} argument.")
|
93
|
+
else
|
94
|
+
self.send("#{field[:name]}=", args[index])
|
95
|
+
end
|
96
|
+
end
|
97
|
+
self.set_input_fields(*(args[self.class.fields.size..-1] || []))
|
98
|
+
end
|
99
|
+
|
100
|
+
def set_input_fields *input_fields
|
101
|
+
self.input_fields = input_fields.map do |field|
|
102
|
+
if field =~ /^[A-Z].*;/
|
103
|
+
self.class.from_string(field)
|
104
|
+
else
|
105
|
+
field
|
106
|
+
end
|
107
|
+
end
|
108
|
+
end
|
109
|
+
|
110
|
+
def output_fields
|
111
|
+
input_fields.map(&:to_s)
|
112
|
+
end
|
113
|
+
|
114
|
+
def to_flat
|
115
|
+
[stream_name].tap do |record|
|
116
|
+
fields.each do |field|
|
117
|
+
value = send(field[:name])
|
118
|
+
record << value.to_s unless value.nil? && field[:optional]
|
119
|
+
end
|
120
|
+
end.concat(output_fields)
|
121
|
+
end
|
122
|
+
|
123
|
+
def to_s
|
124
|
+
to_flat.join(';')
|
125
|
+
end
|
126
|
+
|
127
|
+
end
|
128
|
+
|
129
|
+
end
|
130
|
+
end
|
131
|
+
|