Class: Neuronet::Scale

Inherits:
Object
  • Object
show all
Defined in:
lib/neuronet.rb

Overview

Neuronet::Scale is a class to help scale problems to fit within a network’s “field of view”. Given a list of values, it finds the minimum and maximum values and establishes a mapping to a scaled set of numbers between minus one and one (-1,1).

Direct Known Subclasses

Gaussian

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(factor = 1.0, center = nil, spread = nil) ⇒ Scale

If the value of center is provided, then that value will be used instead of calculating it from the values passed to method set. Likewise, if spread is provided, that value of spread will be used. The attribute @init flags if there is a initiation phase to the calculation of @spread and @center. For Scale, @init is true and the initiation phase calculates the intermediate values @min and @max (the minimum and maximum values in the data set). It’s possible for subclasses of Scale, such as Gaussian, to not have this initiation phase.



329
330
331
332
333
 
# File 'lib/neuronet.rb', line 329

def initialize(factor=1.0,center=nil,spread=nil)
  @factor,@center,@spread = factor,center,spread
  @centered, @spreaded = center.nil?, spread.nil?
  @init = true
end

Instance Attribute Details

#center ⇒ Object

Returns the value of attribute center.



317
318
319
 
# File 'lib/neuronet.rb', line 317

def center
  @center
end

#init=(value) ⇒ Object (writeonly)

Sets the attribute init

Parameters:

  • value —

    the value to set the attribute init to.



318
319
320
 
# File 'lib/neuronet.rb', line 318

def init=(value)
  @init = value
end

#spread ⇒ Object

Returns the value of attribute spread.



317
318
319
 
# File 'lib/neuronet.rb', line 317

def spread
  @spread
end

Instance Method Details

#mapped(inputs) ⇒ Object Also known as: mapped_input, mapped_output 



357
358
359
360
 
# File 'lib/neuronet.rb', line 357

def mapped(inputs)
  factor = 1.0 / (@factor*@spread)
  inputs.map{|value| factor*(value - @center)}
end

#set(inputs) ⇒ Object 



351
352
353
354
355
 
# File 'lib/neuronet.rb', line 351

def set(inputs)
  set_init(inputs)		if @init
  set_center(inputs)	if @centered
  set_spread(inputs)	if @spreaded
end

#set_center(inputs) ⇒ Object

In this case, inputs is unused, but it’s there for the general case.



347
348
349
 
# File 'lib/neuronet.rb', line 347

def set_center(inputs)
  @center = (@max + @min) / 2.0
end

#set_init(inputs) ⇒ Object 



335
336
337
 
# File 'lib/neuronet.rb', line 335

def set_init(inputs)
  @min, @max = inputs.minmax
end

#set_spread(inputs) ⇒ Object

In this case, inputs is unused, but it’s there for the general case.



341
342
343
 
# File 'lib/neuronet.rb', line 341

def set_spread(inputs)
  @spread = (@max - @min) / 2.0
end

#unmapped(outputs) ⇒ Object Also known as: unmapped_input, unmapped_output

Note that it could also unmap inputs, but outputs is typically what’s being transformed back.



366
367
368
369
 
# File 'lib/neuronet.rb', line 366

def unmapped(outputs)
  factor = @factor*@spread
  outputs.map{|value| factor*value + @center}
end