Class: Weka::Core::Instances

Inherits:
Object
  • Object
show all
Includes:
Weka::Concerns::Persistent, Weka::Concerns::Serializable
Defined in:
lib/weka/core/instances.rb

Constant Summary collapse

DEFAULT_RELATION_NAME = 
'Instances'.freeze

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Weka::Concerns::Serializable

#serialize

Methods included from Weka::Concerns::Persistent

included

Constructor Details

#initialize(relation_name: DEFAULT_RELATION_NAME, attributes: []) ⇒ Instances

Returns a new instance of Instances.



43
44
45
46
47
48
 
# File 'lib/weka/core/instances.rb', line 43

def initialize(relation_name: DEFAULT_RELATION_NAME, attributes: [])
  attribute_list = FastVector.new
  attributes.each { |attribute| attribute_list.add_element(attribute) }

  super(relation_name.to_s, attribute_list, 0)
end

Class Method Details

.from_arff(file) ⇒ Object 



20
21
22
 
# File 'lib/weka/core/instances.rb', line 20

def from_arff(file)
  Loader.load_arff(file)
end

.from_c45(file) ⇒ Object

Loads instances based on a given *.names file (holding the attribute values) or a given *.data file (holding the attribute values). The respective other file is loaded from the same directory.

See www.cs.washington.edu/dm/vfml/appendixes/c45.htm for more information about the C4.5 file format.



38
39
40
 
# File 'lib/weka/core/instances.rb', line 38

def from_c45(file)
  Loader.load_c45(file)
end

.from_csv(file) ⇒ Object 



24
25
26
 
# File 'lib/weka/core/instances.rb', line 24

def from_csv(file)
  Loader.load_csv(file)
end

.from_json(file) ⇒ Object 



28
29
30
 
# File 'lib/weka/core/instances.rb', line 28

def from_json(file)
  Loader.load_json(file)
end

Instance Method Details

#add_attributes(&block) ⇒ Object Also known as: with_attributes 



68
69
70
71
 
# File 'lib/weka/core/instances.rb', line 68

def add_attributes(&block)
  instance_eval(&block) if block
  self
end

#add_instance(instance_or_values, weight: 1.0) ⇒ Object

Add new instance

Examples:

Passing Instance

instances.add_instance(instance)

Passing an array of attribute values

attr_values = [attr1_value, attr2_value, attr3_value]
instances.add_instance(attr_values)

Passing a hash of attribute values.

attr_values = { attr1_name: attr1_value, attr2_name: attr2_value }
instances.add_instance(attr_values)

Parameters:

  • instance_or_values (Instance, Array, Hash) —

    the attribute values of the instance to be added. If passing an array, the attribute values must be in the same order as the attributes defined in Instances. If passing a hash, The keys are the names of the attributes and their values are corresponding attributes values.



218
219
220
221
 
# File 'lib/weka/core/instances.rb', line 218

def add_instance(instance_or_values, weight: 1.0)
  instance = instance_from(instance_or_values, weight: weight)
  add(instance)
end

#add_instances(data, weight: 1.0) ⇒ Object 



223
224
225
 
# File 'lib/weka/core/instances.rb', line 223

def add_instances(data, weight: 1.0)
  data.each { |values| add_instance(values, weight: weight) }
end

#apply_filter(filter) ⇒ Object 



247
248
249
 
# File 'lib/weka/core/instances.rb', line 247

def apply_filter(filter)
  filter.filter(self)
end

#apply_filters(*filters) ⇒ Object 



251
252
253
254
255
 
# File 'lib/weka/core/instances.rb', line 251

def apply_filters(*filters)
  filters.inject(self) do |filtered_instances, filter|
    filter.filter(filtered_instances)
  end
end

#attribute_names(include_class_attribute: false) ⇒ Object 



64
65
66
 
# File 'lib/weka/core/instances.rb', line 64

def attribute_names(include_class_attribute: false)
  attributes(include_class_attribute: include_class_attribute).map(&:name)
end

#attributes(include_class_attribute: false) ⇒ Object 



54
55
56
57
58
59
60
61
62
 
# File 'lib/weka/core/instances.rb', line 54

def attributes(include_class_attribute: false)
  attrs = enumerate_attributes.to_a

  if include_class_attribute && class_attribute_defined?
    attrs.insert(class_index, class_attribute)
  end

  attrs
end

#class_attribute ⇒ Object 



188
189
190
 
# File 'lib/weka/core/instances.rb', line 188

def class_attribute
  classAttribute if class_attribute_defined?
end

#class_attribute=(name) ⇒ Object 



174
175
176
177
178
179
180
181
 
# File 'lib/weka/core/instances.rb', line 174

def class_attribute=(name)
  if name.nil?
    reset_class_attribute
  else
    ensure_attribute_defined!(name)
    setClass(attribute_with_name(name))
  end
end

#class_attribute_defined? ⇒ Boolean

Returns:

  • (Boolean) 


196
197
198
 
# File 'lib/weka/core/instances.rb', line 196

def class_attribute_defined?
  class_index >= 0
end

#date(name, format: 'yyyy-MM-dd HH:mm', class_attribute: false) ⇒ Object Also known as: add_date_attribute 



168
169
170
171
172
 
# File 'lib/weka/core/instances.rb', line 168

def date(name, format: 'yyyy-MM-dd HH:mm', class_attribute: false)
  attribute = Attribute.new_date(name, format)
  add_attribute(attribute)
  self.class_attribute = name if class_attribute
end

#each ⇒ Object 



97
98
99
100
101
102
103
 
# File 'lib/weka/core/instances.rb', line 97

def each
  if block_given?
    enumerate_instances.each { |instance| yield(instance) }
  else
    enumerate_instances
  end
end

#each_attribute ⇒ Object 



111
112
113
114
115
116
117
 
# File 'lib/weka/core/instances.rb', line 111

def each_attribute
  if block_given?
    enumerate_attributes.each { |attribute| yield(attribute) }
  else
    enumerate_attributes
  end
end

#each_attribute_with_index ⇒ Object 



119
120
121
122
123
 
# File 'lib/weka/core/instances.rb', line 119

def each_attribute_with_index
  enumerate_attributes.each_with_index do |attribute, index|
    yield(attribute, index) if block_given?
  end
end

#each_with_index ⇒ Object 



105
106
107
108
109
 
# File 'lib/weka/core/instances.rb', line 105

def each_with_index
  enumerate_instances.each_with_index do |instance, index|
    yield(instance, index) if block_given?
  end
end

#has_attribute_type?(type) ⇒ Boolean

Check if the instances has any attribute of the given type

Examples:

Passing String

instances.has_attribute_type?('string')
instances.has_attribute_type?('String')

Passing Symbol

instances.has_attribute_type?(:String)

Passing Integer

instances.has_attribute_type?(Attribute::STRING)

Parameters:

  • type (String, Symbol, Integer) —

    type of the attribute to check String and Symbol argument are converted to corresponding type defined in Weka::Core::Attribute

Returns:

  • (Boolean) 


92
93
94
95
 
# File 'lib/weka/core/instances.rb', line 92

def has_attribute_type?(type)
  type = map_attribute_type(type) unless type.is_a?(Integer)
  check_for_attribute_type(type)
end

#instances ⇒ Object 



50
51
52
 
# File 'lib/weka/core/instances.rb', line 50

def instances
  enumerate_instances.to_a
end

#internal_values_of(values) ⇒ Array, Hash

Retrieve the internal floating point values used to represent

the attributes.

Parameters:

  • values (Array, Hash) —

    the attribute values whose floating point representation should be retrieved.

Returns:

  • (Array, Hash) —

    an array of the internal floating point representation if the input is an Array. Hash otherwise.



235
236
237
238
239
240
241
242
243
244
245
 
# File 'lib/weka/core/instances.rb', line 235

def internal_values_of(values)
  use_hash = values.is_a?(Hash)
  values = attribute_values_from_hash(values) if use_hash

  values = values.map.with_index do |value, index|
    attribute(index).internal_value_of(value)
  end

  values = attribute_values_to_hash(values) if use_hash
  values
end

#merge(*instances) ⇒ Object 



257
258
259
260
261
 
# File 'lib/weka/core/instances.rb', line 257

def merge(*instances)
  instances.inject(self) do |merged_instances, dataset|
    self.class.merge_instances(merged_instances, dataset)
  end
end

#nominal(name, values:, class_attribute: false) ⇒ Object Also known as: add_nominal_attribute 



156
157
158
159
160
 
# File 'lib/weka/core/instances.rb', line 156

def nominal(name, values:, class_attribute: false)
  attribute = Attribute.new_nominal(name, values)
  add_attribute(attribute)
  self.class_attribute = name if class_attribute
end

#numeric(name, class_attribute: false) ⇒ Object Also known as: add_numeric_attribute 



150
151
152
153
154
 
# File 'lib/weka/core/instances.rb', line 150

def numeric(name, class_attribute: false)
  attribute = Attribute.new_numeric(name)
  add_attribute(attribute)
  self.class_attribute = name if class_attribute
end

#reset_class_attribute ⇒ Object 



192
193
194
 
# File 'lib/weka/core/instances.rb', line 192

def reset_class_attribute
  set_class_index(-1)
end

#string(name, class_attribute: false) ⇒ Object Also known as: add_string_attribute 



162
163
164
165
166
 
# File 'lib/weka/core/instances.rb', line 162

def string(name, class_attribute: false)
  attribute = Attribute.new_string(name)
  add_attribute(attribute)
  self.class_attribute = name if class_attribute
end

#to_arff(file) ⇒ Object 



125
126
127
 
# File 'lib/weka/core/instances.rb', line 125

def to_arff(file)
  Saver.save_arff(file: file, instances: self)
end

#to_c45(file) ⇒ Object

Creates a file with the istances’s attribute values and a *.data file with the actual data.

You should choose another file extension than .data (preferably *.names) for the file, else it will just be overwritten with the automatically created *.data file.

See www.cs.washington.edu/dm/vfml/appendixes/c45.htm for more information about the C4.5 file format.



146
147
148
 
# File 'lib/weka/core/instances.rb', line 146

def to_c45(file)
  Saver.save_c45(file: file, instances: self)
end

#to_csv(file) ⇒ Object 



129
130
131
 
# File 'lib/weka/core/instances.rb', line 129

def to_csv(file)
  Saver.save_csv(file: file, instances: self)
end

#to_json(file) ⇒ Object 



133
134
135
 
# File 'lib/weka/core/instances.rb', line 133

def to_json(file)
  Saver.save_json(file: file, instances: self)
end

#to_m ⇒ Matrix

Get the all instances’s values as Matrix.

Returns:

  • (Matrix) —

    a Matrix holding the instance’s values as rows.



266
267
268
 
# File 'lib/weka/core/instances.rb', line 266

def to_m
  Matrix[*instances.map(&:values)]
end