def test_l1l2_multi_task_squared_hinge_loss(mult_dense_train_data):
mult_dense, mult_target = mult_dense_train_data
Y = LabelBinarizer(neg_label=-1).fit_transform(mult_target)
clf = CDClassifier(penalty="l1/l2",
loss="squared_hinge",
multiclass=False,
max_iter=20,
C=5.0,
random_state=0)
clf.fit(mult_dense, mult_target)
df = clf.decision_function(mult_dense)
np.testing.assert_array_almost_equal(clf.errors_.T, 1 - Y * df)
np.testing.assert_almost_equal(clf.score(mult_dense, mult_target), 0.8633,
3)
nz = np.sum(clf.coef_ != 0)
assert nz == 300
clf = CDClassifier(penalty="l1/l2",
loss="squared_hinge",
multiclass=False,
max_iter=20,
C=0.05,
random_state=0)
clf.fit(mult_dense, mult_target)
np.testing.assert_almost_equal(clf.score(mult_dense, mult_target), 0.8266,
3)
nz = np.sum(clf.coef_ != 0)
assert nz == 231
def test_l1l2_multiclass_log_loss(data, request):
X, y = request.getfixturevalue(data)
clf = CDClassifier(penalty="l1/l2",
loss="log",
multiclass=True,
max_steps=30,
max_iter=5,
C=1.0,
random_state=0)
clf.fit(X, y)
np.testing.assert_almost_equal(clf.score(X, y), 0.8766, 3)
df = clf.decision_function(X)
sel = np.array([df[i, int(y[i])] for i in range(df.shape[0])])
df -= sel[:, np.newaxis]
df = np.exp(df)
np.testing.assert_array_almost_equal(clf.errors_, df.T)
for i in range(X.shape[0]):
np.testing.assert_almost_equal(clf.errors_[y[i], i], 1.0)
nz = np.sum(clf.coef_ != 0)
assert nz == 297
clf = CDClassifier(penalty="l1/l2",
loss="log",
multiclass=True,
max_steps=30,
max_iter=5,
C=0.3,
random_state=0)
clf.fit(X, y)
np.testing.assert_almost_equal(clf.score(X, y), 0.8566, 3)
nz = np.sum(clf.coef_ != 0)
assert nz == 213
assert nz % 3 == 0 # should be a multiple of n_classes
def test_l1l2_multiclass_squared_hinge_loss(data, request):
X, y = request.getfixturevalue(data)
clf = CDClassifier(penalty="l1/l2",
loss="squared_hinge",
multiclass=True,
max_iter=20,
C=1.0,
random_state=0)
clf.fit(X, y)
np.testing.assert_almost_equal(clf.score(X, y), 0.913, 3)
df = clf.decision_function(X)
n_samples, n_vectors = df.shape
diff = np.zeros_like(clf.errors_)
for i in range(n_samples):
for k in range(n_vectors):
diff[k, i] = 1 - (df[i, y[i]] - df[i, k])
np.testing.assert_array_almost_equal(clf.errors_, diff)
assert np.sum(clf.coef_ != 0) == 300
clf = CDClassifier(penalty="l1/l2",
loss="squared_hinge",
multiclass=True,
max_iter=20,
C=0.05,
random_state=0)
clf.fit(X, y)
np.testing.assert_almost_equal(clf.score(X, y), 0.83, 3)
nz = np.sum(clf.coef_ != 0)
assert nz == 207
assert nz % 3 == 0 # should be a multiple of n_classes
def test_l1l2_multiclass_squared_hinge_loss_no_linesearch():
data = mult_csc
clf = CDClassifier(penalty="l1/l2",
loss="squared_hinge",
multiclass=True,
shrinking=False,
selection="uniform",
max_steps=0,
max_iter=200,
C=1.0,
random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.9166, 3)
df = clf.decision_function(data)
n_samples, n_vectors = df.shape
diff = np.zeros_like(clf.errors_)
for i in xrange(n_samples):
for k in xrange(n_vectors):
diff[k, i] = 1 - (df[i, mult_target[i]] - df[i, k])
assert_array_almost_equal(clf.errors_, diff)
assert_equal(np.sum(clf.coef_ != 0), 300)
clf = CDClassifier(penalty="l1/l2",
loss="squared_hinge",
multiclass=True,
max_iter=20,
C=0.05,
random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.83, 3)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 207)
assert_true(nz % 3 == 0) # should be a multiple of n_classes
def test_l1l2_multiclass_log_loss():
for data in (mult_dense, mult_csc):
clf = CDClassifier(penalty="l1/l2",
loss="log",
multiclass=True,
max_steps=30,
max_iter=5,
C=1.0,
random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.8766, 3)
df = clf.decision_function(data)
sel = np.array(
[df[i, int(mult_target[i])] for i in xrange(df.shape[0])])
df -= sel[:, np.newaxis]
df = np.exp(df)
assert_array_almost_equal(clf.errors_, df.T)
for i in xrange(data.shape[0]):
assert_almost_equal(clf.errors_[mult_target[i], i], 1.0)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 297)
clf = CDClassifier(penalty="l1/l2",
loss="log",
multiclass=True,
max_steps=30,
max_iter=5,
C=0.3,
random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.8566, 3)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 213)
assert_true(nz % 3 == 0) # should be a multiple of n_classes
def test_l1l2_multi_task_squared_hinge_loss():
Y = LabelBinarizer(neg_label=-1).fit_transform(mult_target)
clf = CDClassifier(penalty="l1/l2", loss="squared_hinge",
multiclass=False,
max_iter=20, C=5.0, random_state=0)
clf.fit(mult_dense, mult_target)
df = clf.decision_function(mult_dense)
assert_array_almost_equal(clf.errors_.T, 1 - Y * df)
assert_almost_equal(clf.score(mult_dense, mult_target), 0.8633, 3)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 300)
clf = CDClassifier(penalty="l1/l2", loss="squared_hinge",
multiclass=False,
max_iter=20, C=0.05, random_state=0)
clf.fit(mult_dense, mult_target)
assert_almost_equal(clf.score(mult_dense, mult_target), 0.8266, 3)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 231)
def test_l1l2_multiclass_log_loss():
for data in (mult_dense, mult_csc):
clf = CDClassifier(penalty="l1/l2", loss="log", multiclass=True,
max_steps=30, max_iter=5, C=1.0, random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.8766, 3)
df = clf.decision_function(data)
sel = np.array([df[i, int(mult_target[i])] for i in xrange(df.shape[0])])
df -= sel[:, np.newaxis]
df = np.exp(df)
assert_array_almost_equal(clf.errors_, df.T)
for i in xrange(data.shape[0]):
assert_almost_equal(clf.errors_[mult_target[i], i], 1.0)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 297)
clf = CDClassifier(penalty="l1/l2", loss="log", multiclass=True,
max_steps=30, max_iter=5, C=0.3, random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.8566, 3)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 213)
assert_true(nz % 3 == 0) # should be a multiple of n_classes
def test_l1l2_multiclass_squared_hinge_loss_no_linesearch():
data = mult_csc
clf = CDClassifier(penalty="l1/l2", loss="squared_hinge",
multiclass=True, shrinking=False, selection="uniform",
max_steps=0, max_iter=200, C=1.0, random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.9166, 3)
df = clf.decision_function(data)
n_samples, n_vectors = df.shape
diff = np.zeros_like(clf.errors_)
for i in xrange(n_samples):
for k in xrange(n_vectors):
diff[k, i] = 1 - (df[i, mult_target[i]] - df[i, k])
assert_array_almost_equal(clf.errors_, diff)
assert_equal(np.sum(clf.coef_ != 0), 300)
clf = CDClassifier(penalty="l1/l2", loss="squared_hinge",
multiclass=True,
max_iter=20, C=0.05, random_state=0)
clf.fit(data, mult_target)
assert_almost_equal(clf.score(data, mult_target), 0.83, 3)
nz = np.sum(clf.coef_ != 0)
assert_equal(nz, 207)
assert_true(nz % 3 == 0) # should be a multiple of n_classes
