def test_load_lfw_pairs():
if not os.path.exists(os.path.join(get_data_home(), 'lfw_home')):
raise SkipTest
lfw_pairs_train = load_lfw_pairs(subset='train')
# this dataset is used for training supervised face verification models,
# this is a binary classification task
top_classes = ['Different persons', 'Same person']
assert_array_equal(lfw_pairs_train.target_names, top_classes)
# default slice is a rectangular shape around the face, removing
# most of the background, for each of the 2 face pictures
assert_equal(lfw_pairs_train.data.shape, (2200, 2, 62, 47))
# the ordering is respecting the metadata text file of the official LFW
# tasks
assert_equal(lfw_pairs_train.target.shape, (2200, ))
assert_array_equal(lfw_pairs_train.target[:5], [1, 1, 1, 1, 1])
assert_array_equal(lfw_pairs_train.target[-5:], [0, 0, 0, 0, 0])
# as for the people loader it is also possible to load the color channels
# in the last dimension
lfw_pairs_train = load_lfw_pairs(subset='train', color=True)
assert_equal(lfw_pairs_train.data.shape, (2200, 2, 62, 47, 3))
# the data also has a test development set and a 10-fold CV dataset for
# final evaluation
lfw_pairs_test = load_lfw_pairs(subset='test')
assert_equal(lfw_pairs_test.data.shape, (1000, 2, 62, 47))
lfw_pairs_10_folds = load_lfw_pairs(subset='10_folds')
assert_equal(lfw_pairs_10_folds.data.shape, (6000, 2, 62, 47))
def test_load_fake_lfw_pairs():
lfw_pairs_train = load_lfw_pairs(data_home=SCIKIT_LEARN_DATA)
# The data is croped around the center as a rectangular bounding box
# arounthe the face. Colors are converted to gray levels:
assert_equal(lfw_pairs_train.data.shape, (10, 2, 62, 47))
# the target is whether the person is the same or not
assert_array_equal(lfw_pairs_train.target, [1, 1, 1, 1, 1, 0, 0, 0, 0, 0])
# names of the persons can be found using the target_names array
expected_classes = ['Different persons', 'Same person']
assert_array_equal(lfw_pairs_train.target_names, expected_classes)
# It is possible to ask for the original data without any croping or color
# conversion
lfw_pairs_train = load_lfw_pairs(data_home=SCIKIT_LEARN_DATA,
resize=None,
slice_=None,
color=True)
assert_equal(lfw_pairs_train.data.shape, (10, 2, 250, 250, 3))
# the ids and class names are the same as previously
assert_array_equal(lfw_pairs_train.target, [1, 1, 1, 1, 1, 0, 0, 0, 0, 0])
assert_array_equal(lfw_pairs_train.target_names, expected_classes)
def test_load_lfw_pairs():
if not os.path.exists(os.path.join(get_data_home(), 'lfw_home')):
raise SkipTest
lfw_pairs_train = load_lfw_pairs(subset='train')
# this dataset is used for training supervised face verification models,
# this is a binary classification task
top_classes = ['Different persons', 'Same person']
assert_array_equal(lfw_pairs_train.target_names, top_classes)
# default slice is a rectangular shape around the face, removing
# most of the background, for each of the 2 face pictures
assert_equal(lfw_pairs_train.data.shape, (2200, 2, 62, 47))
# the ordering is respecting the metadata text file of the official LFW
# tasks
assert_equal(lfw_pairs_train.target.shape, (2200,))
assert_array_equal(lfw_pairs_train.target[:5], [1, 1, 1, 1, 1])
assert_array_equal(lfw_pairs_train.target[-5:], [0, 0, 0, 0, 0])
# as for the people loader it is also possible to load the color channels
# in the last dimension
lfw_pairs_train = load_lfw_pairs(subset='train', color=True)
assert_equal(lfw_pairs_train.data.shape, (2200, 2, 62, 47, 3))
# the data also has a test development set and a 10-fold CV dataset for
# final evaluation
lfw_pairs_test = load_lfw_pairs(subset='test')
assert_equal(lfw_pairs_test.data.shape, (1000, 2, 62, 47))
lfw_pairs_10_folds = load_lfw_pairs(subset='10_folds')
assert_equal(lfw_pairs_10_folds.data.shape, (6000, 2, 62, 47))
def test_load_fake_lfw_pairs():
lfw_pairs_train = load_lfw_pairs(data_home=SCIKIT_LEARN_DATA)
# The data is croped around the center as a rectangular bounding box
# arounthe the face. Colors are converted to gray levels:
assert_equal(lfw_pairs_train.data.shape, (10, 2, 62, 47))
# the target is whether the person is the same or not
assert_array_equal(lfw_pairs_train.target, [1, 1, 1, 1, 1, 0, 0, 0, 0, 0])
# names of the persons can be found using the target_names array
expected_classes = ['Different persons', 'Same person']
assert_array_equal(lfw_pairs_train.target_names, expected_classes)
# It is possible to ask for the original data without any croping or color
# conversion
lfw_pairs_train = load_lfw_pairs(data_home=SCIKIT_LEARN_DATA,
resize=None, slice_=None, color=True)
assert_equal(lfw_pairs_train.data.shape, (10, 2, 250, 250, 3))
# the ids and class names are the same as previously
assert_array_equal(lfw_pairs_train.target, [1, 1, 1, 1, 1, 0, 0, 0, 0, 0])
assert_array_equal(lfw_pairs_train.target_names, expected_classes)
def test_load_empty_lfw_pairs():
lfw_people = load_lfw_pairs(data_home=SCIKIT_LEARN_EMPTY_DATA)
def test_load_empty_lfw_pairs():
lfw_people = load_lfw_pairs(data_home=SCIKIT_LEARN_EMPTY_DATA)