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

#*****************************************************************************80
#
## newswire uses keras to classify newswires from the Reuters dataset.
#
#  Discussion:
#
#    Each newswire is to be assigned to one of 46 separate classes.
#
#  Licensing:
#
#    This code is distributed under the MIT license.
#
#  Modified:
#
#    22 April 2019
#
#  Author:
#
#    The original text is by Francois Chollet;
#    some modifications by John Burkardt.
#
#  Reference:
#
#    Francois Chollet,
#    Deep Learning with Python,
#    Manning, 2018,
#    ISBN: 9781617294433.
#
  import keras
  import platform
  
  print ( '' )
  print ( 'newswire:' )
  print ( '  python version: %s' % ( platform.python_version ( ) ) )
  print ( '  keras version: %s' % ( keras.__version__ ) )
  print ( '  Neural network to classify the Reuters newswire data.' )
#
#  Import and load the dataset.
#
  from keras.datasets import reuters

  ( train_data, train_labels ), ( test_data, test_labels ) = reuters.load_data ( num_words = 10000 )
#
#  Display the size of the training and testing items.
#
  print ( '' )
  print ( '  Number of training data items = %d' % ( len ( train_data ) ) )
  print ( '  Number of test data items = %d' % ( len ( test_data ) ) )
#
#  Exhibit the contents of one item of the training data:
#
  print ( '  Sample train_data[10]:' )
  print ( train_data[10] )
#
#  Show how to convert the numeric data back into text.
#
  word_index = reuters.get_word_index ( )
  reverse_word_index = dict ( [ ( value, key ) for ( key, value ) in word_index.items ( ) ] )
  decoded_newswire = ''.join ( [ reverse_word_index.get ( i - 3, '?' ) for i in train_data[0] ] )
#
#  The label that is associated with a data item is the topic index, between 0 and 45.
#
  print ( '  Sample train_label[10]:' )
  print ( train_labels[10] )
#
#  Vectorize the data.
#
  import numpy as np

  def vectorize_sequences ( sequences, dimension = 10000 ):
    results = np.zeros ( ( len ( sequences ), dimension ) )
    for i, sequence in enumerate ( sequences ):
      results[i,sequence] = 1.0
    return results

  x_train = vectorize_sequences ( train_data )
  x_test  = vectorize_sequences ( test_data  )
#
#  Vectorize the labels using one-hot encoding.
#  (A vector of 0's with a single 1).
#
  from keras.utils.np_utils import to_categorical

  one_hot_train_labels = to_categorical ( train_labels )
  one_hot_test_labels  = to_categorical ( test_labels  )
#
#  Define the network architecture.
#
  from keras import models
  from keras import layers

  model = models.Sequential()
  model.add ( layers.Dense ( 64, activation = 'relu', input_shape = ( 10000, ) ) )
  model.add ( layers.Dense ( 64, activation = 'relu' ) )
  model.add ( layers.Dense ( 46, activation = 'softmax' ) )
#
#  Compile the network.
#
  model.compile ( \
    optimizer = 'rmsprop', \
    loss = 'categorical_crossentropy', \
    metrics = [ 'accuracy' ] )
#
#  Select 1000 data items for validation.
#
  x_val = x_train[:1000]
  partial_x_train = x_train[1000:]
  y_val = one_hot_train_labels[:1000]
  partial_y_train = one_hot_train_labels[1000:]
#
#  Train the network for 20 epochs
#
  history = model.fit ( \
    partial_x_train, \
    partial_y_train, \
    epochs = 20, \
    batch_size = 512, \
    validation_data = ( x_val, y_val ) )
#
#  Display the loss curve.
#
  import matplotlib.pyplot as plt

  loss = history.history['loss']
  val_loss = history.history['val_loss']

  epochs = range ( 1, len ( loss ) + 1 )

  plt.plot ( epochs, loss,     'bo', label = 'Training loss' )
  plt.plot ( epochs, val_loss, 'b',  label = 'Validation loss' )
  plt.title ( 'Training and validation loss' )
  plt.xlabel ( 'Epochs' )
  plt.ylabel ( 'Loss' )
  plt.legend ( )
  filename = 'newswire_loss.png'
  plt.savefig ( filename )
  print ( '' )
  print ( '  Graphics saved in "%s"' % ( filename ) )
  plt.show ( )
#
#  Display the accuracy curve.
#
  plt.clf ( )

  acc = history.history['acc']
  val_acc = history.history['val_acc']

  plt.plot ( epochs, acc,     'bo', label = 'Training accuracy' )
  plt.plot ( epochs, val_acc, 'b',  label = 'Validation accuracy' )
  plt.title ( 'Training and validation accuracy' )
  plt.xlabel ( 'Epochs' )
  plt.ylabel ( 'Loss' )
  plt.legend ( )
  filename = 'newswire_accuracy.png'
  plt.savefig ( filename )
  print ( '' )
  print ( '  Graphics saved in "%s"' % ( filename ) )
  plt.show ( )
#
#  Terminate.
#
  print ( '' )
  print ( 'newswire:' )
  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 ( )
  newswire ( )
  timestamp ( )