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

#*****************************************************************************80
#
## faces_classify_nmf() uses nonnegative matrix factorization (nmf) on face data.
#
#  Licensing:
#
#    This code is distributed under the MIT license.
#
#  Modified:
#
#    21 September 2023
#
#  Author:
#
#    Andreas Mueller, Sarah Guido.
#    This version by John Burkardt.
#
#  Reference:
#
#    Andreas Mueller, Sarah Guido,
#    Introduction to Machine Learning with Python,
#    OReilly, 2017,
#    ISBN: 978-1-449-36941-5
#
  from sklearn.datasets import fetch_lfw_people
  from sklearn.decomposition import NMF
  from sklearn.model_selection import train_test_split
  from sklearn.neighbors import KNeighborsClassifier
  import matplotlib.pyplot as plt
  import mglearn
  import numpy as np
  import platform
  import sklearn

  print ( '' )
  print ( 'faces_classify_nmf():' )
  print ( '  Python version: ' + platform.python_version ( ) )
  print ( '  scikit-learn version: '+ sklearn.__version__ )
  print ( '  Match a face to an item in the faces dataset.' )
  print ( '  Use nonnegative matrix factorization (nmf).' )
  print ( '' )
#
#  Show the results of NMF on the two-dimensional toy data:
#
  mglearn.plots.plot_nmf_illustration ( )
  filename = 'toy_nmf.png'
  plt.savefig ( filename )
  print ( '  Graphics saved as "' + filename + '"' )
#
#  Generate the dataset.
#
  print ( '  Generate the dataset' )
  people = fetch_lfw_people ( min_faces_per_person = 20, resize = 0.7 )
  image_shape = people.images[0].shape
#
#  Trim the dataset so no more than 50 images of any person.
#
  mask = np.zeros ( people.target.shape, dtype = bool )
  for target in np.unique ( people.target ):
    mask[np.where ( people.target == target) [0][:50]] = 1 
  X_people = people.data[mask]
  y_people = people.target[mask]
#
#  Scale the grayscale values to be between 0 and 1.
#
  X_people = X_people / 255.0
#
#  Split the data.
#
  X_train, X_test, y_train, y_test = train_test_split ( 
    X_people, y_people, stratify = y_people, random_state = 0 )
#
#  Show how the number of components affects the quality of the reconstruction.
#
  if ( False ):
    mglearn.plots.plot_nmf_faces ( X_train, X_test, image_shape )
    filename = 'faces_reconstruction_nmf.png'
    plt.savefig ( filename )
    print ( '  Graphics saved as "' + filename + '"' )
#
#  Extract first 15 nmf omponents.
#
  nmf = NMF ( n_components = 15, random_state = 0, max_iter = 6000 )
  nmf.fit ( X_train )
  X_train_nmf = nmf.transform ( X_train )
  X_test_nmf = nmf.transform ( X_test )
#
#  Plot nmf components.
#
  plt.clf ( )
  fig, axes = plt.subplots ( 3, 5, figsize = ( 15, 12 ),
    subplot_kw = { 'xticks':(), 'yticks':() } )
  for i, ( component, ax ) in enumerate ( zip ( nmf.components_, axes.ravel() ) ):
    ax.imshow ( component.reshape ( image_shape ) )
    ax.set_title ( "Component {}".format ( ( i+1) ) )

  filename = 'nmf_components.png'
  plt.savefig ( filename )
  print ( "  Graphics saved as '" + filename + "'" )
#
#  Look at images for which these components are strong.
#  Sort by component 3, plot first 10 images.
#
  compn = 3
  inds = np.argsort ( X_train_nmf[:,compn])[::-1]
  fig, axes = plt.subplots ( 2, 5, figsize = ( 15, 8 ),
    subplot_kw = { 'xticks':(), 'yticks':() } )
  fig.suptitle ( "Large component 3" )
  for i, ( ind, ax ) in enumerate ( zip ( inds, axes.ravel() ) ):
    ax.imshow ( X_train[ind].reshape(image_shape) )

  filename = 'nmf_component3.png'
  plt.savefig ( filename )
  print ( "  Graphics saved as '" + filename + "'" )
#
#  Sort by component 7, plot first 10 images.
#
  compn = 7
  inds = np.argsort ( X_train_nmf[:,compn])[::-1]
  fig, axes = plt.subplots ( 2, 5, figsize = ( 15, 8 ),
    subplot_kw = { 'xticks':(), 'yticks':() } )
  fig.suptitle ( "Large component 7" )
  for i, ( ind, ax ) in enumerate ( zip ( inds, axes.ravel() ) ):
    ax.imshow ( X_train[ind].reshape(image_shape) )

  filename = 'nmf_component7.png'
  plt.savefig ( filename )
  print ( "  Graphics saved as '" + filename + "'" )
#
#  Terminate.
#
  print ( '' )
  print ( 'faces_classify_nmf():' )
  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:
#
#    21 August 2019
#
#  Author:
#
#    John Burkardt
#
  import time

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

  return

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