def process(self):
"""!
@brief Performs cluster analysis in line with rules of K-Medoids algorithm.
@return (kmedoids) Returns itself (K-Medoids instance).
@remark Results of clustering can be obtained using corresponding get methods.
@see get_clusters()
@see get_medoids()
"""
if self.__ccore is True:
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__clusters, self.__medoid_indexes = wrapper.kmedoids(self.__pointer_data, self.__medoid_indexes, self.__tolerance, self.__itermax, ccore_metric.get_pointer(), self.__data_type)
else:
changes = float('inf')
iterations = 0
while changes > self.__tolerance and iterations < self.__itermax:
self.__clusters = self.__update_clusters()
update_medoid_indexes = self.__update_medoids()
changes = max([self.__distance_calculator(self.__medoid_indexes[index], update_medoid_indexes[index]) for index in range(len(update_medoid_indexes))])
self.__medoid_indexes = update_medoid_indexes
iterations += 1
return self
def __initialize_by_ccore(self):
"""!
@brief Performs processing using C/C++ pyclustering library.
"""
ccore_metric = metric_wrapper.create_instance(self.__metric)
return pam_build_wrapper.pam_build(self.__data, self.__amount, ccore_metric.get_pointer(), self.__data_type)
def testBuildGowerDistanceFromMetricWithNumpyData(self):
metric = distance_metric(type_metric.GOWER,
data=numpy.array([[-3.0, -3.0], [-4.0, -3.0],
[-4.5, -3.0], [-5.0,
-3.0]]))
ccore_metric = metric_wrapper.create_instance(metric)
self.assertEqual(0.5, ccore_metric([-3.0, -3.0], [-5.0, -3.0]))
def process(self):
"""!
@brief Performs cluster analysis in line with rules of K-Medoids algorithm.
@remark Results of clustering can be obtained using corresponding get methods.
@see get_clusters()
@see get_medoids()
"""
if self.__ccore is True:
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__clusters = wrapper.kmedoids(self.__pointer_data, self.__medoid_indexes, self.__tolerance, ccore_metric.get_pointer(), self.__data_type)
self.__medoid_indexes = self.__update_medoids()
else:
changes = float('inf')
stop_condition = self.__tolerance
while changes > stop_condition:
self.__clusters = self.__update_clusters()
update_medoid_indexes = self.__update_medoids()
changes = max([self.__distance_calculator(self.__medoid_indexes[index], update_medoid_indexes[index]) for index in range(len(update_medoid_indexes))])
self.__medoid_indexes = update_medoid_indexes
def __process_by_ccore(self):
"""!
@brief Performs processing using CCORE (C/C++ part of pyclustering library).
"""
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__score = wrapper.silhoeutte(self.__data, self.__clusters, ccore_metric.get_pointer(), self.__data_type)
def process(self):
"""!
@brief Performs cluster analysis in line with rules of K-Medians algorithm.
@remark Results of clustering can be obtained using corresponding get methods.
@see get_clusters()
@see get_medians()
"""
if self.__ccore is True:
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__clusters, self.__medians = wrapper.kmedians(self.__pointer_data, self.__medians, self.__tolerance, ccore_metric.get_pointer())
else:
changes = float('inf')
# Check for dimension
if len(self.__pointer_data[0]) != len(self.__medians[0]):
raise NameError('Dimension of the input data and dimension of the initial medians must be equal.')
while changes > self.__tolerance:
self.__clusters = self.__update_clusters()
updated_centers = self.__update_medians()
changes = max([self.__metric(self.__medians[index], updated_centers[index]) for index in range(len(updated_centers))])
self.__medians = updated_centers
def testBuildGowerDistanceFromMetricWithNumpyData(self):
metric = distance_metric(type_metric.GOWER, data=numpy.array([[-3.0, -3.0], [-4.0, -3.0], [-4.5, -3.0], [-5.0, -3.0]]))
ccore_metric = metric_wrapper.create_instance(metric)
self.assertEqual(0.5, ccore_metric([-3.0, -3.0], [-5.0, -3.0]))
# TODO: doesn't work for some platforms.
#def testUserDefinedMetric(self):
# user_metric = lambda p1, p2 : p1[0] + p2[0];
# metric_instance = metric_wrapper(type_metric.USER_DEFINED, [], user_metric);
# assertion.eq(2.0, metric_instance([0.0, 0.0], [2.0, 0.0]));
# assertion.eq(4.0, metric_instance([3.0, 2.0], [1.0, 5.0]));
def __process_by_ccore(self):
"""!
@brief Performs cluster analysis using CCORE (C/C++ part of pyclustering library).
"""
ccore_metric = metric_wrapper.create_instance(self.__metric)
result = wrapper.xmeans(self.__pointer_data, self.__centers, self.__kmax, self.__tolerance, self.__criterion,
self.__alpha, self.__beta, self.__repeat, self.__random_state,
ccore_metric.get_pointer())
self.__clusters = result[0]
self.__centers = result[1]
self.__total_wce = result[2][0]
def __process_by_ccore(self):
"""!
@brief Performs cluster analysis using CCORE (C/C++ part of pyclustering library).
"""
ccore_metric = metric_wrapper.create_instance(self.__metric)
results = wrapper.kmeans(self.__pointer_data, self.__centers, self.__tolerance, self.__itermax, (self.__observer is not None), ccore_metric.get_pointer())
self.__clusters = results[0]
self.__centers = results[1]
if self.__observer is not None:
self.__observer.set_evolution_clusters(results[2])
self.__observer.set_evolution_centers(results[3])
self.__total_wce = results[4][0]
def __process_by_ccore(self):
"""!
@brief Performs cluster analysis using CCORE (C/C++ part of pyclustering library).
"""
ccore_metric = metric_wrapper.create_instance(self.__metric)
results = wrapper.kmeans(self.__pointer_data, self.__centers, self.__tolerance, self.__itermax, (self.__observer is not None), ccore_metric.get_pointer())
self.__clusters = results[0]
self.__centers = results[1]
if self.__observer is not None:
self.__observer.set_evolution_clusters(results[2])
self.__observer.set_evolution_centers(results[3])
self.__total_wce = results[4][0]
def process(self):
"""!
@brief Performs cluster analysis in line with rules of K-Medoids algorithm.
@return (kmedoids) Returns itself (K-Medoids instance).
@remark Results of clustering can be obtained using corresponding get methods.
@see get_clusters()
@see get_medoids()
"""
if self.__ccore is True:
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__clusters, self.__medoid_indexes, self.__iterations, self.__total_deviation = kmedoids_wrapper.kmedoids(self.__pointer_data, self.__medoid_indexes, self.__tolerance, self.__itermax, ccore_metric.get_pointer(), self.__data_type)
# TODO: calculate it on C++ side as well
for index_cluster in range(len(self.__clusters)):
for index_point in self.__clusters[index_cluster]:
self.__labels[index_point] = index_cluster
else:
changes = float('inf')
previous_deviation, self.__total_deviation = float('inf'), 0
self.__iterations = 0
if self.__itermax > 0:
self.__total_deviation = self.__update_clusters()
while (changes > self.__tolerance) and (self.__iterations < self.__itermax):
self.__iterations += 1
swap_cost = self.__swap_medoids()
if swap_cost != float('inf'):
previous_deviation = self.__total_deviation
self.__total_deviation = self.__update_clusters()
changes = previous_deviation - self.__total_deviation
else:
break
self.__erase_empty_clusters()
return self
def process(self):
"""!
@brief Performs cluster analysis in line with rules of K-Medoids algorithm.
@return (kmedoids) Returns itself (K-Medoids instance).
@remark Results of clustering can be obtained using corresponding get methods.
@see get_clusters()
@see get_medoids()
"""
if self.__ccore is True:
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__clusters, self.__medoid_indexes = wrapper.kmedoids(
self.__pointer_data, self.__medoid_indexes, self.__tolerance,
self.__itermax, ccore_metric.get_pointer(), self.__data_type)
else:
changes = float('inf')
previous_deviation, current_deviation = float('inf'), float('inf')
iterations = 0
if self.__itermax > 0:
current_deviation = self.__update_clusters()
while (changes > self.__tolerance) and (iterations <
self.__itermax):
swap_cost = self.__swap_medoids()
if swap_cost != float('inf'):
previous_deviation = current_deviation
current_deviation = self.__update_clusters()
changes = previous_deviation - current_deviation
else:
break
iterations += 1
self.__erase_empty_clusters()
return self
def process(self):
"""!
@brief Performs cluster analysis in line with rules of K-Medians algorithm.
@return (kmedians) Returns itself (K-Medians instance).
@remark Results of clustering can be obtained using corresponding get methods.
@see get_clusters()
@see get_medians()
"""
if self.__ccore is True:
ccore_metric = metric_wrapper.create_instance(self.__metric)
self.__clusters, self.__medians = wrapper.kmedians(self.__pointer_data, self.__medians, self.__tolerance, self.__itermax, ccore_metric.get_pointer())
else:
changes = float('inf')
# Check for dimension
if len(self.__pointer_data[0]) != len(self.__medians[0]):
raise NameError('Dimension of the input data and dimension of the initial medians must be equal.')
iterations = 0
while changes > self.__tolerance and iterations < self.__itermax:
self.__clusters = self.__update_clusters()
updated_centers = self.__update_medians()
changes = max([self.__metric(self.__medians[index], updated_centers[index]) for index in range(len(updated_centers))])
self.__medians = updated_centers
iterations += 1
return self
def __process_by_ccore(self):
ccore_metric = metric_wrapper.create_instance(self._metric);
self._clusters, self._representatives = bsas_wrapper(self._data, self._amount, self._threshold, ccore_metric.get_pointer());
def testBuildGowerDistanceFromMetricWithNumpyMaxRange(self):
metric = distance_metric(type_metric.GOWER,
max_range=numpy.array([2.0, 0.0]))
ccore_metric = metric_wrapper.create_instance(metric)
self.assertEqual(0.5, ccore_metric([-3.0, -3.0], [-5.0, -3.0]))
def testBuildGowerDistanceFromMetricWithMaxRange(self):
metric = distance_metric(type_metric.GOWER, max_range=[2.0, 0.0])
ccore_metric = metric_wrapper.create_instance(metric)
assertion.eq(0.5, ccore_metric([-3.0, -3.0], [-5.0, -3.0]))
def __process_by_ccore(self):
ccore_metric = metric_wrapper.create_instance(self._metric)
self._clusters, self._representatives = mbsas_wrapper(
self._data, self._amount, self._threshold,
ccore_metric.get_pointer())
def testBuildGowerDistanceFromMetricWithData(self):
metric = distance_metric(type_metric.GOWER,
data=[[-3.0, -3.0], [-4.0, -3.0],
[-4.5, -3.0], [-5.0, -3.0]])
ccore_metric = metric_wrapper.create_instance(metric)
assertion.eq(0.5, ccore_metric([-3.0, -3.0], [-5.0, -3.0]))
