#! /usr/bin/env python3
#
def regression_1d ( ):

#*****************************************************************************80
#
## regression_1d uses keras to solve a simple 1d regression case.
#
#  Licensing:
#
#    This code is distributed under the MIT license.
#
#  Modified:
#
#    25 April 2019
#
#  Author:
#
#    John Burkardt.
#
#  Reference:
#
#    Francois Chollet,
#    Deep Learning with Python,
#    Manning, 2018,
#    ISBN: 9781617294433.
#
  import keras
  import matplotlib.pyplot as plt
  import numpy as np
  import platform
  
  print ( '' )
  print ( 'regression_1d:' )
  print ( '  python version: %s' % ( platform.python_version ( ) ) )
  print ( '  keras version: %s' % ( keras.__version__ ) )
  print ( '  Neural network to solve a 1d regression problem.' )
  print ( '  The data is read from external files.' )
#
#  Load the datasets from external files.
#
  train = np.loadtxt ( 'regression_1d_train.txt' )
  print ( train.shape )
  train_data = train[:,0]
  train_targets = train[:,1]
  train_num = len ( train_data )

  test = np.loadtxt ( 'regression_1d_test.txt' )
  print ( test.shape )
  test_data = test[:,0]
  test_targets = test[:,1]
  test_num = len ( test_data )
#
#  Exhibit the contents of one item of the training data:
#
  print ( train_data[0:3], train_data[-3:] )
  print ( train_targets[0:3], train_targets[-3:] )
#
#  Normalize the data.
#
# mean = train_data.mean ( axis = 0 )
# train_data = train_data - mean
# std = train_data.std ( axis = 0 )
# train_data = train_data / std

# test_data = test_data - mean
# test_data = test_data / std

  from keras import models
  from keras import layers

  model = models.Sequential()
  model.add ( layers.Dense ( input_dim = 1, output_dim = 1 ) )
  model.compile ( optimizer = 'sgd', loss = 'mse' )

  print ( '' )
  print ( 'Training:' )
  print ( '' )

  for step in range ( 101 ):
    cost = model.train_on_batch ( train_data, train_targets )
    if ( step % 10 == 0 ):
      print ( 'step %d: training cost = %g' % ( step, cost ) )
#
#  Test
#
  print ( '' )
  print ( 'Testing:' )
  print ( '' )
  cost = model.evaluate ( test_data, test_targets, batch_size = test_num )
  print ( 'Test cost:', cost )
  W, b = model.layers[0].get_weights ( )
  print ( 'Weights = ', W ) 
  print ( 'Bias = ', b )
#
#  Plot the data.
#
  xmin = -1.0
  xmax = 2.0
  plt.plot ( [xmin, xmax ], [ b + W[0] * xmin, b + W[0] * xmax], 'r-' )
  plt.plot ( train_data, train_targets, 'bo' )
  plt.plot ( test_data, test_targets, 'ro' )
  plt.grid ( True )
  plt.xlabel ( '<--- X --->', fontsize = 16 )
  plt.ylabel ( '<--- Y --->', fontsize = 16 )
  plt.title ( 'Data: train (blue), test (red)', fontsize = 16 )
  filename = 'regression_1d.png'
  plt.savefig ( filename )
  print ( '' )
  print ( '  Graphics saved as "%s"' % ( filename ) )
  plt.show ( )
#
#  Terminate.
#
  print ( '' )
  print ( 'regression_1d:' )
  print ( '  Normal end of execution.' )
  return

def timestamp ( ):

#*****************************************************************************80
#
## TIMESTAMP prints the date as a timestamp.
#
#  Licensing:
#
#    This code is distributed under the MIT license. 
#
#  Modified:
#
#    06 April 2013
#
#  Author:
#
#    John Burkardt
#
#  Parameters:
#
#    None
#
  import time

  t = time.time ( )
  print ( time.ctime ( t ) )

  return None

if ( __name__ == '__main__' ):
  timestamp ( )
  regression_1d ( )
  timestamp ( )