def test_rbf_kernel(setup):
rs = np.random.RandomState(0)
raw_X = rs.rand(10, 4)
raw_Y = rs.rand(11, 4)
r = rbf_kernel(raw_X, raw_Y)
result = r.to_numpy()
expected = sklearn_rbf_kernel(raw_X, raw_Y)
np.testing.assert_almost_equal(result, expected)
def testRbfKernel(self):
rs = np.random.RandomState(0)
raw_X = rs.rand(10, 4)
raw_Y = rs.rand(11, 4)
r = rbf_kernel(raw_X, raw_Y)
result = r.execute()
expected = sklearn_rbf_kernel(raw_X, raw_Y)
np.testing.assert_almost_equal(result, expected)
def setUp(self) -> None:
self.estimators = [(LabelPropagation, {
'kernel': 'rbf'
}), (LabelPropagation, {
'kernel': 'knn',
'n_neighbors': 2
}),
(LabelPropagation, {
'kernel':
lambda x, y: rbf_kernel(x, y, gamma=20)
})]
def setUp(self) -> None:
self.session = new_session().as_default()
self._old_executor = self.session._sess._executor
self.executor = self.session._sess._executor = \
ExecutorForTest('numpy', storage=self.session._sess._context)
self.estimators = [(LabelPropagation, {
'kernel': 'rbf'
}), (LabelPropagation, {
'kernel': 'knn',
'n_neighbors': 2
}),
(LabelPropagation, {
'kernel':
lambda x, y: rbf_kernel(x, y, gamma=20)
})]
from sklearn.model_selection import train_test_split
from sklearn.exceptions import ConvergenceWarning
from sklearn.utils._testing import assert_no_warnings, assert_warns
from mars import tensor as mt
from mars.learn.metrics.pairwise import rbf_kernel
from mars.learn.neighbors import NearestNeighbors
from mars.learn.semi_supervised import LabelPropagation
estimators = [(LabelPropagation, {
'kernel': 'rbf'
}), (LabelPropagation, {
'kernel': 'knn',
'n_neighbors': 2
}), (LabelPropagation, {
'kernel': lambda x, y: rbf_kernel(x, y, gamma=20)
})]
@pytest.mark.parametrize('estimator, parameters', estimators)
def test_fit_transduction(setup, estimator, parameters):
samples = [[1., 0.], [0., 2.], [1., 3.]]
labels = [0, 1, -1]
clf = estimator(**parameters).fit(samples, labels)
assert clf.transduction_[2].fetch() == 1
@pytest.mark.parametrize('estimator, parameters', estimators)
def test_distribution(setup, estimator, parameters):
samples = [[1., 0.], [0., 1.], [1., 1.]]
labels = [0, 1, -1]
