RubyでTensorFlowを使う with PyCall ref: http://qiita.com/mix_dvd/items/ab46ac5e6bf48c45a4e5

file0.sh

$ gem install --pre pycall

file1.sh

$ gem install --pre matplotlib

file1.txt

y = w \times x + b

file2.txt

y = w \times x + b

file3.txt

$ ruby tf.rb
0   -0.075883001    0.61687899
20  0.031335868 0.34092095
40  0.082482144 0.31043994
60  0.095530763 0.30266351
80  0.098859809 0.30067953
100 0.099709123 0.30017337
120 0.099925794 0.30004424
140 0.099981084 0.30001128
160 0.099995196 0.30000287
180 0.099998787 0.30000073
200 0.099999689 0.30000019

file4.txt

$ ruby tf.rb
0   -0.075883001    0.61687899
20  0.031335868 0.34092095
40  0.082482144 0.31043994
60  0.095530763 0.30266351
80  0.098859809 0.30067953
100 0.099709123 0.30017337
120 0.099925794 0.30004424
140 0.099981084 0.30001128
160 0.099995196 0.30000287
180 0.099998787 0.30000073
200 0.099999689 0.30000019

file5.txt

# 略

# Fit the line.
w_log = []
b_log = []

(0..200).each do |step|
  sess.run.(train)
  #if step % 20 == 0 then
  #  puts "#{step}\t#{sess.run.(w)[0]}\t#{sess.run.(b)[0]}"
  #end
  w_log << sess.run.(w)[0]
  b_log << sess.run.(b)[0]
end
# Learns best fit is W: [0.1], b: [0.3]

pyimport 'matplotlib.pyplot', as: 'plt'
require 'matplotlib/iruby'
Matplotlib::IRuby.activate

plt.plot.(w_log)
plt.plot.(b_log)
plt.show.()

tf.rb

require 'pycall/import'
include PyCall::Import

pyimport 'numpy', as: 'np'
pyimport 'tensorflow', as: 'tf'

# Create 100 phony x, y data points in NumPy, y = x * 0.1 + 0.3
x_data = np.random.rand.(100).astype.(np.float32)
y_data = x_data * 0.1 + 0.3

# Try to find values for W and b that compute y_data = W * x_data + b
# (We know that W should be 0.1 and b 0.3, but Tensorflow will
# figure that out for us.)
w = tf.Variable.(tf.random_uniform.([1], -1.0, 1.0))
b = tf.Variable.(tf.zeros.([1]))
y = w * x_data + b

# Minimize the mean squared errors.
loss = tf.reduce_mean.(tf.square.(y - y_data))
optimizer = tf.train.GradientDescentOptimizer.(0.5)
train = optimizer.minimize.(loss)

# Before starting, initialize the variables. We will 'run' this first.
init = tf.initialize_all_variables.()

# Launch the graph.
sess = tf.Session.()
sess.run.(init)

# Fit the line.
(0..200).each do |step|
  sess.run.(train)
  if step % 20 == 0 then
    puts "#{step}\t#{sess.run.(w)[0]}\t#{sess.run.(b)[0]}"
  end
end
# Learns best fit is W: [0.1], b: [0.3]