iris.cr

require "csv"
require "../crystal-fann"

# https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data
data = CSV.parse(File.read("./iris.data"))
puts "Size: #{data.size}" # => 150
data.shuffle!

classes = {
  "Iris-setosa"     => [1.0, 0.0, 0.0],
  "Iris-versicolor" => [0.0, 1.0, 0.0],
  "Iris-virginica"  => [0.0, 0.0, 1.0],
}

# Split train and test data
train, test = data[0..100], data[100..150]

ann = Fann::Network::Standard.new(4, [10], 3)

train_X = train.map { |i| [i[0], i[1], i[2], i[3]].map(&.to_f64) }
train_y = train.map { |i| classes[i[4]] }

test_X = test.map { |i| [i[0], i[1], i[2], i[3]].map(&.to_f64) }
test_y = test.map { |i| classes[i[4]] }

10_000.times do
  train_X.size.times do |i|
    ann.train_single(train_X[i], train_y[i])
  end
end

def result(y)
  max = y.index(y.max)
  Array.new(3, 0).map_with_index do |_, i|
    next 1 if i == max
    0
  end
end

errors = 0
total = test_X.size
total.times do |i|
  next errors += 1 if result(ann.run(test_X[i])) != test_y[i]
  puts test_y[i]
end

puts "result: #{1 - (errors.to_f / total).round(2)} (#{errors}/#{total})"

ann.close