def makeSmartNData(features, targets, isregression, distance=False):
"""
Some comment..
:param features:
:param targets:
:param isregression:
:return:
"""
num_classes = targets.shape[1]
targetscopy = targets.copy()
nonbinarized = _inverse_binarize_multiclass(
targetscopy, [i for i in range(num_classes)])
digit_indices = [
np.where(nonbinarized == i)[0] for i in range(num_classes)
]
classcounter = {
key: {
"counter": 1,
"max": len(digit_indices[key])
}
for key in range(0, num_classes)
}
combineddata = np.zeros((features.shape[0], (2 * features.shape[1])))
targets2 = np.zeros((targets.shape[0], targets.shape[1]))
combineddata[:, 0:features.shape[1]] = features
sim_target = np.zeros((targets.shape[0], 1))
positive = True
targetind = 0
for i in range(0, targets.shape[0]):
if positive:
# this time pick other part of pair from same class (positive pair)
newdata, targetind = getSameClass(features, nonbinarized[i],
digit_indices, num_classes,
classcounter)
combineddata[i, features.shape[1]:features.shape[1] * 2] = newdata
if distance:
sim_target[i] = 0
else:
sim_target[i] = 1
positive = False
else:
# pick the other part of the pair from another
# class (negative pair)
newdata, targetind = getOtherClass(features, nonbinarized[i],
digit_indices, num_classes,
classcounter)
combineddata[i, features.shape[1]:features.shape[1] * 2] = newdata
if distance:
sim_target[i] = 1
else:
sim_target[i] = 0
positive = True
targets2[i] = targets[targetind]
return combineddata, sim_target, targets.copy(), \
targets2
def check_binarized_results(y, classes, pos_label, neg_label, expected):
for sparse_output in [True, False]:
if ((pos_label == 0 or neg_label != 0) and sparse_output):
assert_raises(ValueError,
label_binarize,
y,
classes,
neg_label=neg_label,
pos_label=pos_label,
sparse_output=sparse_output)
continue
# check label_binarize
binarized = label_binarize(y,
classes,
neg_label=neg_label,
pos_label=pos_label,
sparse_output=sparse_output)
assert_array_equal(toarray(binarized), expected)
assert_equal(issparse(binarized), sparse_output)
# check inverse
y_type = type_of_target(y)
if y_type == "multiclass":
inversed = _inverse_binarize_multiclass(binarized, classes=classes)
else:
inversed = _inverse_binarize_thresholding(
binarized,
output_type=y_type,
classes=classes,
threshold=((neg_label + pos_label) / 2.))
assert_array_equal(toarray(inversed), toarray(y))
# Check label binarizer
lb = LabelBinarizer(neg_label=neg_label,
pos_label=pos_label,
sparse_output=sparse_output)
binarized = lb.fit_transform(y)
assert_array_equal(toarray(binarized), expected)
assert_equal(issparse(binarized), sparse_output)
inverse_output = lb.inverse_transform(binarized)
assert_array_equal(toarray(inverse_output), toarray(y))
assert_equal(issparse(inverse_output), issparse(y))
def check_binarized_results(y, classes, pos_label, neg_label, expected):
for sparse_output in [True, False]:
if ((pos_label == 0 or neg_label != 0) and sparse_output):
assert_raises(ValueError, label_binarize, y, classes,
neg_label=neg_label, pos_label=pos_label,
sparse_output=sparse_output)
continue
# check label_binarize
binarized = label_binarize(y, classes, neg_label=neg_label,
pos_label=pos_label,
sparse_output=sparse_output)
assert_array_equal(toarray(binarized), expected)
assert_equal(issparse(binarized), sparse_output)
# check inverse
y_type = type_of_target(y)
if y_type == "multiclass":
inversed = _inverse_binarize_multiclass(binarized, classes=classes)
else:
inversed = _inverse_binarize_thresholding(binarized,
output_type=y_type,
classes=classes,
threshold=((neg_label +
pos_label) /
2.))
assert_array_equal(toarray(inversed), toarray(y))
# Check label binarizer
lb = LabelBinarizer(neg_label=neg_label, pos_label=pos_label,
sparse_output=sparse_output)
binarized = lb.fit_transform(y)
assert_array_equal(toarray(binarized), expected)
assert_equal(issparse(binarized), sparse_output)
inverse_output = lb.inverse_transform(binarized)
assert_array_equal(toarray(inverse_output), toarray(y))
assert_equal(issparse(inverse_output), issparse(y))
def test_inverse_binarize_multiclass():
got = _inverse_binarize_multiclass(csr_matrix([[0, 1, 0],
[-1, 0, -1],
[0, 0, 0]]),
np.arange(3))
assert_array_equal(got, np.array([1, 1, 0]))
def test_inverse_binarize_multiclass():
got = _inverse_binarize_multiclass(csr_matrix([[0, 1, 0],
[-1, 0, -1],
[0, 0, 0]]),
np.arange(3))
assert_array_equal(got, np.array([1, 1, 0]))
