def test_convergence_simple_example(capsys): # LMNN should converge on this simple example, which it did not with # this issue: https://github.com/metric-learn/metric-learn/issues/88 X, y = make_classification(random_state=0) lmnn = python_LMNN(verbose=True) lmnn.fit(X, y) out, _ = capsys.readouterr() assert "LMNN converged with objective" in out
def test_convergence_simple_example(capsys):
# LMNN should converge on this simple example, which it did not with
# this issue: https://github.com/metric-learn/metric-learn/issues/88
X, y = make_classification(random_state=0)
lmnn = python_LMNN(verbose=True)
lmnn.fit(X, y)
out, _ = capsys.readouterr()
assert "LMNN converged with objective" in out
def _process_inputs(self, X, labels):
num_pts = X.shape[0]
assert len(labels) == num_pts
unique_labels, self.label_inds = np.unique(labels, return_inverse=True)
self.labels = np.arange(len(unique_labels))
self.X = X
# self.L = np.eye(X.shape[1])
model_lmnn = lmnn.python_LMNN(k=4, min_iter=50, learn_rate=1e-7)
model_lmnn.fit(X=X[:100], labels=labels[:100])
self.L = model_lmnn.transformer()
def test_no_twice_same_objective(capsys):
# test that the objective function never has twice the same value
# see https://github.com/metric-learn/metric-learn/issues/88
X, y = make_classification(random_state=0)
lmnn = python_LMNN(verbose=True)
lmnn.fit(X, y)
out, _ = capsys.readouterr()
lines = re.split("\n+", out)
# we get only objectives from each line:
# the regexp matches a float that follows an integer (the iteration
# number), and which is followed by a (signed) float (delta obj). It
# matches for instance:
# 3 **1113.7665747189938** -3.182774197440267 46431.0200999999999998e-06
objectives = [re.search("\d* (?:(\d*.\d*))[ | -]\d*.\d*", s)
for s in lines]
objectives = [match.group(1) for match in objectives if match is not None]
# we remove the last element because it can be equal to the penultimate
# if the last gradient update is null
assert len(objectives[:-1]) == len(set(objectives[:-1]))
def test_no_twice_same_objective(capsys):
# test that the objective function never has twice the same value
# see https://github.com/metric-learn/metric-learn/issues/88
X, y = make_classification(random_state=0)
lmnn = python_LMNN(verbose=True)
lmnn.fit(X, y)
out, _ = capsys.readouterr()
lines = re.split("\n+", out)
# we get only objectives from each line:
# the regexp matches a float that follows an integer (the iteration
# number), and which is followed by a (signed) float (delta obj). It
# matches for instance:
# 3 **1113.7665747189938** -3.182774197440267 46431.0200999999999998e-06
objectives = [
re.search("\d* (?:(\d*.\d*))[ | -]\d*.\d*", s) for s in lines
]
objectives = [match.group(1) for match in objectives if match is not None]
# we remove the last element because it can be equal to the penultimate
# if the last gradient update is null
assert len(objectives[:-1]) == len(set(objectives[:-1]))
train_sets_my_model_100 = my_model_100.transform(data[:110])
test_sets_my_model_100 = my_model_100.transform(data[110:])
knn_my_mode_100 = neighbors.KNeighborsClassifier(n_neighbors=3)
knn_my_mode_100.fit(train_sets_my_model_100, target[:110])
print(my_model_100.metric())
print("knn_before:", knn_before.score(data[110:], target[110:]))
print("knn_after:", knn_my_mode_100.score(test_sets_my_model_100,
target[110:]))
print("my_mode_100 improve:",(knn_my_mode_100.score(test_sets_my_model_100,target[110:])-
knn_before.score(data[110:],target[110:]))/\
knn_before.score(data[110:],target[110:])*100,"%")
'''
LMNN
'''
lmnn = lmnn.python_LMNN(k=3, learn_rate=1e-6)
#lmnn = My_model.python_My_ML(learn_rate=1,alpha=1,max_iter=10)
lmnn.fit(data[:110], target[:110])
#print(lmnn.metric())
train_sets_lmnn = lmnn.transform(data[:110])
test_sets_lmnn = lmnn.transform(data[110:])
knn_lmnn = neighbors.KNeighborsClassifier(n_neighbors=3)
knn_lmnn.fit(train_sets_lmnn, target[:110])
print(lmnn.metric())
print("knn_before:", knn_before.score(data[110:], target[110:]))
print("knn_after:", knn_lmnn.score(test_sets_lmnn, target[110:]))
print("lmnn improve:",(knn_lmnn.score(test_sets_lmnn,target[110:])-
knn_before.score(data[110:],target[110:]))/\
knn_before.score(data[110:],target[110:])*100,"%")
'Covariance': metric_learn.Covariance(),
'ITML_Supervised': metric_learn.ITML_Supervised(num_constraints=200),
'LFDA': metric_learn.LFDA(k=2, dim=2),
'LMNN': metric_learn.LMNN(k=5, learn_rate=1e-6, verbose=False),
'LSML_Supervised': metric_learn.LSML_Supervised(num_constraints=200),
'MLKR': metric_learn.MLKR(),
'NCA': metric_learn.NCA(max_iter=700, learning_rate=0.01, num_dims=2),
'RCA_Supervised': metric_learn.RCA_Supervised(dim=2, num_chunks=30,
chunk_size=2),
'SDML_Supervised': metric_learn.SDML_Supervised(num_constraints=1500),
}
try:
from metric_learn.lmnn import python_LMNN
if python_LMNN is not metric_learn.LMNN:
CLASSES['python_LMNN'] = python_LMNN(k=5, learn_rate=1e-6, verbose=False)
except ImportError:
pass
class IrisDataset(object):
params = [sorted(CLASSES)]
param_names = ['alg']
def setup(self, alg):
iris_data = load_iris()
self.iris_points = iris_data['data']
self.iris_labels = iris_data['target']
def time_fit(self, alg):
np.random.seed(5555)
'Covariance': metric_learn.Covariance(),
'ITML_Supervised': metric_learn.ITML_Supervised(num_constraints=200),
'LFDA': metric_learn.LFDA(k=2, dim=2),
'LMNN': metric_learn.LMNN(k=5, learn_rate=1e-6, verbose=False),
'LSML_Supervised': metric_learn.LSML_Supervised(num_constraints=200),
'MLKR': metric_learn.MLKR(),
'NCA': metric_learn.NCA(max_iter=700, num_dims=2),
'RCA_Supervised': metric_learn.RCA_Supervised(dim=2, num_chunks=30,
chunk_size=2),
'SDML_Supervised': metric_learn.SDML_Supervised(num_constraints=1500),
}
try:
from metric_learn.lmnn import python_LMNN
if python_LMNN is not metric_learn.LMNN:
CLASSES['python_LMNN'] = python_LMNN(k=5, learn_rate=1e-6, verbose=False)
except ImportError:
pass
class IrisDataset(object):
params = [sorted(CLASSES)]
param_names = ['alg']
def setup(self, alg):
iris_data = load_iris()
self.iris_points = iris_data['data']
self.iris_labels = iris_data['target']
def time_fit(self, alg):
np.random.seed(5555)