from graph_measures.features_algorithms.vertices.louvain import LouvainCalculator
from graph_measures.features_algorithms.vertices.motifs import nth_nodes_motif
from graph_measures.features_algorithms.vertices.page_rank import PageRankCalculator
from graph_measures.features_infra.feature_calculators import FeatureMeta, FeatureCalculator
from graph_measures.features_algorithms.vertices.neighbor_nodes_histogram import nth_neighbor_calculator
# from graph_measures.features_algorithms.accelerated_graph_features.attractor_basin import AttractorBasinCalculator
# from graph_measures.features_algorithms.accelerated_graph_features.bfs_moments import BfsMomentsCalculator
# from graph_measures.features_algorithms.accelerated_graph_features.flow import FlowCalculator
# from graph_measures.features_algorithms.accelerated_graph_features.k_core import KCoreCalculator
# from graph_measures.features_algorithms.accelerated_graph_features.motifs import nth_nodes_motif
# from graph_measures.features_algorithms.accelerated_graph_features.page_rank import PageRankCalculator
NODE_FEATURES = {
# Passed
"attractor_basin":
FeatureMeta(AttractorBasinCalculator, {"ab"}),
"average_neighbor_degree":
FeatureMeta(AverageNeighborDegreeCalculator, {"avg_nd"}),
"betweenness_centrality":
FeatureMeta(BetweennessCentralityCalculator, {"betweenness"}),
"bfs_moments":
FeatureMeta(BfsMomentsCalculator, {"bfs"}),
# Didn't pass - but no logic
"closeness_centrality":
FeatureMeta(ClosenessCentralityCalculator, {"closeness"}),
# "communicability_betweenness_centrality": FeatureMeta(CommunicabilityBetweennessCentralityCalculator,
# {"communicability"}),
# Passed
"eccentricity":
for node in self._gnx}
max_b_u = float(max(b_u.values()))
for node in self._gnx:
# the delta determines whether this node is to be considered
if (b_u[node] / max_b_u) <= self._threshold:
self._features[node] = 0
continue
udists = undirected_dists[node]
dists = directed_dists[node]
# getting coordinated values from two dictionaries with the same keys
# saving the data as np.array type
num, denom = map(np.array, zip(*((udists[n], dists[n]) for n in dists)))
num = num[denom != 0]
denom = denom[denom != 0]
self._features[node] = np.sum(num / denom) / float(b_u[node])
feature_entry = {
"flow": FeatureMeta(FlowCalculator, {}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(FlowCalculator, is_max_connected=True)
import networkx as nx
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class BetweennessCentralityCalculator(NodeFeatureCalculator):
def __init__(self, *args, normalized=False, **kwargs):
super(BetweennessCentralityCalculator, self).__init__(*args, **kwargs)
self._is_normalized = normalized
def _calculate(self, include: set, is_regression=False):
self._features = nx.betweenness_centrality(
self._gnx, normalized=self._is_normalized)
def is_relevant(self):
return True
feature_entry = {
"betweenness_centrality":
FeatureMeta(BetweennessCentralityCalculator, {"betweenness"}),
}
if __name__ == "__main__":
from measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(BetweennessCentralityCalculator,
is_max_connected=True)
d = r
return laplacian, y, tol, r, d
@staticmethod
def _initialize_vars_from_laplacian_matrix1(g):
# creating laplacian matrix
w = g + g.T
d = np.diag(sum(w))
l = d - w
_id = np.sum(g, 0)
od = np.sum(g, 1)
# initialize_vars
b = np.subtract((np.array([od])).T, (np.array([_id])).T)
tol = 0.001
n = np.size(g, 1)
y = np.random.rand(n, 1)
y = np.subtract(y, (1 / n) * sum(y))
k = np.dot(l, y)
r = np.subtract(b, k)
d = r
return l, y, tol, r, d
feature_entry = {
"hierarchy_energy": FeatureMeta(HierarchyEnergyCalculator, {"hierarchy"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(HierarchyEnergyCalculator, is_max_connected=True)
[[cur_feature[r_type][x] for x in range(self._num_classes)]
for r_type in self._relation_types]).flatten()
# def _to_ndarray(self):
# mx = np.matrix([self._get_feature(node) for node in self._nodes()])
# return mx.astype(np.float32)
def nth_neighbor_calculator(order, **kwargs):
# Kwargs are used to pass the labels to consider argument
return partial(NthNeighborNodeHistogramCalculator, order, **kwargs)
feature_entry = {
"first_neighbor_histogram":
FeatureMeta(nth_neighbor_calculator(1), {"fnh", "first_neighbor"}),
"second_neighbor_histogram":
FeatureMeta(nth_neighbor_calculator(2), {"snh", "second_neighbor"}),
}
def build_sample_graph(edges, colors, color_list):
dg = nx.DiGraph(labels=color_list) # list(set(colors.values())))
dg.add_edges_from(edges)
for n in dg:
dg.node[n]['label'] = colors[n]
return dg
def test_neighbor_histogram():
all_colors = ['red', 'blue', 'green', 'yellow']
def _edge_based_degree_directed(self):
for edge in self._gnx.edges():
e1_feature = np.array(self._general_c.feature(edge[0]))
e2_feature = np.array(self._general_c.feature(edge[1]))
self._features[edge] = np.concatenate([
(e1_feature - e2_feature).astype(np.float32), # sub out-in
np.mean([e1_feature, e2_feature],
axis=1).astype(np.float32), # mean out-in
])
def _edge_based_degree_undirected(self):
for edge in self._gnx.edges():
e1_feature = self._general_c.feature(edge[0])
e2_feature = self._general_c.feature(edge[1])
self._features[edge] = [
float(e1_feature[0]) - e2_feature[0], # sub
(float(e1_feature[0]) + e2_feature[0]) / 2 # mean
]
feature_entry = {
"Edge_degree_based_calculator":
FeatureMeta(EdgeDegreeBasedCalculator, {"e_degree"}),
}
if __name__ == 'main':
pass
import sys
sys.path.append(os.path.abspath('.'))
sys.path.append(os.path.abspath('..'))
sys.path.append(os.path.abspath('../..'))
sys.path.append(os.path.abspath('../../..'))
sys.path.append(os.path.abspath('src'))
sys.path.append(os.path.abspath('src/accelerated_graph_features'))
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
from graph_measures.features_algorithms.accelerated_graph_features.src import node_page_rank
class PageRankCalculator(NodeFeatureCalculator):
def __init__(self, *args, alpha=0.9, **kwargs):
super(PageRankCalculator, self).__init__(*args, **kwargs)
self._alpha = alpha
def is_relevant(self):
# Undirected graphs will be converted to a directed
# graph with two directed edges for each undirected edge.
return True
def _calculate(self, include: set):
self._features = node_page_rank(self._gnx, dumping=self._alpha)
feature_entry = {
"page_rank": FeatureMeta(PageRankCalculator, {"pr"}),
}
import os
import sys
sys.path.append(os.path.abspath('.'))
sys.path.append(os.path.abspath('..'))
sys.path.append(os.path.abspath('../..'))
sys.path.append(os.path.abspath('../../..'))
sys.path.append(os.path.abspath('src'))
sys.path.append(os.path.abspath('src/accelerated_graph_features'))
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
from graph_measures.features_algorithms.accelerated_graph_features.src import k_core
class KCoreCalculator(NodeFeatureCalculator):
def is_relevant(self):
return True
def _calculate(self, include: set):
self._features = k_core(self._gnx)
feature_entry = {
"k_core": FeatureMeta(KCoreCalculator, {"kc"}),
}
import networkx as nx
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class CommunicabilityBetweennessCentralityCalculator(NodeFeatureCalculator):
def _calculate(self, include: set, is_regression=False):
self._features = nx.communicability_betweenness_centrality(self._gnx)
def is_relevant(self):
return not self._gnx.is_directed()
feature_entry = {
"communicability_betweenness_centrality": FeatureMeta(CommunicabilityBetweennessCentralityCalculator,
{"communicability"}),
}
if __name__ == "__main__":
from measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(CommunicabilityBetweennessCentralityCalculator, is_max_connected=True)
def _active_learning(self, eps=0.05, batch_size=5, epochs=200, out_prog=True, clear_model=False, out_interval=25,
**params):
if self._opt not in Available_Options:
print("option {} is not supported".format(self._opt))
return
# preliminary calculation for exploitation
neighbors_matrix = None
if self._opt in {'region_entropy'} or params.get('representation_region', False):
gn = GraphNeighbors(self._graph)
neighbors_matrix = gn.neighbors(second_order=params.get('region_second_order', False),
self_node=params.get('region_include_self', False),
with_orders=False, only_out=params.get('region_only_out', False))
if params.get('region_weights', None):
if params['region_weights'] == 'out':
neighbors_degree = np.array(self._graph.out_degree(self._nodes_order))[:, 1]
elif params['region_weights'] == 'in':
neighbors_degree = np.array(self._graph.in_degree(self._nodes_order))[:, 1]
else:
neighbors_degree = np.array(self._graph.degree(self._nodes_order))[:, 1]
neighbors_degree = neighbors_degree.astype(float)
for x in neighbors_matrix:
for j in range(len(x) - 1, -1, -1):
if neighbors_degree[x[j]] == 0:
del x[j]
if params.get('region_opposite_weights', False):
weights = neighbors_degree
else:
weights = 1 / neighbors_degree
weights[weights == np.inf] = 0
else:
weights = None
if self._opt in {'centrality', 'Chang', 'geo_cent', 'APR'}:
page_rank_feature = {"page_rank": FeatureMeta(PageRankCalculator, {"pr"})}
page_rank_matrix = self._gl.get_features(page_rank_feature)
norm_page_rank = normalize(page_rank_matrix, axis=0, norm='max').reshape(-1)
if self._opt == 'APR':
adj_matrix = nx.adjacency_matrix(self._graph, nodelist=self._nodes_order).todense()
np.fill_diagonal(adj_matrix, 0)
adj_matrix = adj_matrix / adj_matrix.sum(axis=1)
adj_matrix = np.nan_to_num(adj_matrix)
adj_matrix = adj_matrix.T
if torch.cuda.is_available():
page_rank_matrix = torch.Tensor(page_rank_matrix).to(self._device)
adj_matrix = torch.Tensor(adj_matrix).to(self._device)
if self._opt in {'rep_dist', 'Chang'}:
inv_cov = None
features_matrix = None
if params.get('representation_vectors', 'model') in {'topology', 'both'}:
rep_features = CHOSEN_FEATURES.copy()
if not params.get('representation_motif4', False):
rep_features.pop('motif4')
features_matrix = self._gl.get_features(rep_features, print_time=True)
# replace all nan values of attractor basin to 100
features_matrix[np.isnan(features_matrix)] = 100
if params.get('representation_measure', 'mahalanobis') == 'mahalanobis' and \
params['representation_vectors'] == 'topology':
inv_cov = np.linalg.pinv(np.cov(features_matrix.T))
if self._opt in {'geo_dist', 'geo_cent'}:
graph_dists = dict(weighted.all_pairs_dijkstra_path_length(self._graph, self._num_nodes, weight='weight'))
ancestors = [nx.ancestors(self._graph, node) for node in sorted(self._graph)]
descendants = [nx.descendants(self._graph, node) for node in sorted(self._graph)]
if self._opt == 'k_truss':
k_truss_score = k_truss(self._graph)
if self._opt == 'feature':
if params.get('feature', 'attractor_basin') == 'attractor_basin':
chosen_feature = {"attractor_basin": FeatureMeta(AttractorBasinCalculator, {"ab"})}
feature_score = self._gl.get_features(chosen_feature)
# initializing parameters for plots
if out_prog:
results = {0: self._init(n_nodes=1)}
prog_intervals = out_interval
prog_iter = 1
cur_interval = 1
else:
self._init()
results = {}
while self._num_revealed < self._stop_cond:
# making output for plot
if out_prog and self._num_revealed >= cur_interval/prog_intervals*self._stop_cond:
res = self._train_and_eval(epochs, iterations=prog_iter, clear_model=clear_model)
percent_revealed = round(self._num_revealed / self._num_nodes * 100, 2)
results[percent_revealed] = res
val = list(res.values())[0]
self._model_runner.data_logger.info(self.eval_method, percent_revealed,
val["loss"], val["acc"], val["mic_f1"], val["mac_f1"])
cur_interval = np.floor(self._num_revealed/self._stop_cond*prog_intervals) + 1
elif self._opt in {'entropy', 'region_entropy', 'rep_dist', 'Chang'}:
self._train_and_eval(epochs, clear_model=clear_model, verbose=0)
# eps greedy algorithm choosing nodes to reveal
rand = np.random.uniform(0, 1)
# Explore
if rand < eps or self._opt == 'random':
idx = self._explore(batch_size=batch_size, balance=params.get('balance', False))
# Exploit
else:
# evaluate the samples
if self._opt in {'rep_dist', 'Chang'}:
outlier_score = self._representation_distance(vectors_type=params.get('representation_vectors',
'model'),
dist_measure=params.get('representation_measure',
'mahalanobis'),
features_matrix=features_matrix, inv_cov=inv_cov,
average_distance=params.get('representation_average',
False),
region=params.get('representation_region', False),
neighbors_matrix=neighbors_matrix)
if self._opt in {'entropy', 'Chang'}:
test_probs = self._model_runner.predict_proba(self._models).cpu()
if params.get('margin', False):
test_probs.sort()
entropies = test_probs[:, -1] - test_probs[:, -2]
else:
entropies = entropy(test_probs.t())
if self._opt in {'region_entropy'}:
region_entropies = self._region_entropies(neighbors_matrix=neighbors_matrix, weights=weights,
average_entropy=params.get('region_average_entropy',
False),
prefer_large_regions=params.get('region_prefer_large',
False),
margin=params.get('margin', False))
if self._opt in {'geo_dist', 'geo_cent'}:
dist_to_train = self._geodesic_distance(graph_dists, ancestors, descendants,
in_out=params.get('geo_in_out', 'both'))
if self._opt == 'APR':
relative_pr = self._relative_page_rank(page_rank_matrix, adj_matrix,
only_known=params.get('apr_only_known', False))
# build scores vector according to the evaluation method
if self._opt == 'entropy':
scores = entropies
elif self._opt == 'region_entropy':
scores = region_entropies
elif self._opt == 'rep_dist':
scores = outlier_score
elif self._opt == 'geo_dist':
scores = dist_to_train
elif self._opt == 'centrality':
scores = np.asarray([page_rank_matrix[x] for x in self._pool]).reshape(-1)
elif self._opt == 'k_truss':
scores = np.asarray([k_truss_score[x] for x in self._pool]).reshape(-1)
elif self._opt == 'feature':
scores = np.asarray([feature_score[x] for x in self._pool]).reshape(-1)
elif self._opt == 'APR':
scores = relative_pr.flat
elif self._opt == 'Chang':
c3 = max(0.7 * (1 - self._num_revealed/self._stop_cond), 0)
c1 = c2 = (1 - c3) / 2
density = normalize(np.nan_to_num(outlier_score).reshape(1, -1)).reshape(-1)
norm_entropy = normalize(entropies.reshape(1, -1)).reshape(-1)
centrality = np.asarray([norm_page_rank[x] for x in self._pool]).reshape(-1)
scores = c1 * norm_entropy + c2 * density + c3 * centrality
elif self._opt == 'geo_cent':
c1 = 0.7
c2 = 0.3
g_dists = normalize(dist_to_train.reshape(1, -1), norm='max').reshape(-1)
centrality = np.vstack(norm_page_rank[x] for x in self._pool).reshape(-1)
scores = c1 * g_dists + c2 * centrality
if params.get('margin', False):
# choose the lowest scores nodes
idx = np.argpartition(scores, batch_size)[:batch_size]
else:
# choose the best nodes
idx = np.argpartition(scores, -batch_size)[-batch_size:]
self._reveal(idx)
# evaluate the model
res = self._train_and_eval(epochs, clear_model=clear_model)
percent_revealed = round(self._num_revealed / self._num_nodes * 100, 2)
results[percent_revealed] = res
val = list(res.values())[0]
self._model_runner.data_logger.info(self.eval_method, percent_revealed,
val["loss"], val["acc"], val["mic_f1"], val["mac_f1"])
return results
denominator = sum((dist / avg_out[m]) * (self._alpha**(-m))
for m, dist in out_dist.items())
if 0 != denominator:
numerator = sum((dist / avg_in[m]) * (self._alpha**(-m))
for m, dist in in_dist.items())
self._features[node] = numerator / denominator
@staticmethod
def _calculate_average_per_dist(num_nodes, count_dist):
# rearrange the details in "count_dist" to be with unique distance in the array "all_dist_count"
all_dist_count = {}
for counter in count_dist.values():
for dist, occurrences in counter.items():
all_dist_count[dist] = all_dist_count.get(dist,
0) + occurrences
# calculating for each distance the average
return {
dist: float(count) / num_nodes
for dist, count in all_dist_count.items()
}
feature_entry = {
"attractor_basin": FeatureMeta(AttractorBasinCalculator, {"ab"}),
}
if __name__ == "__main__":
from measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(AttractorBasinCalculator, is_max_connected=True)
motif_number: 0
for motif_number in self._all_motifs
}
self._features[edge][motif_num] += 1
def nth_nodes_motif(motif_level):
return partial(MotifsNodeCalculator, level=motif_level)
def nth_edges_motif(motif_level):
return partial(MotifsNodeCalculator, level=motif_level)
feature_node_entry = {
"motif3": FeatureMeta(nth_nodes_motif(3), {"m3"}),
"motif4": FeatureMeta(nth_nodes_motif(4), {"m4"}),
}
feature_edge_entry = {
"motif3_edge": FeatureMeta(nth_edges_motif(3), {"me3"}),
"motif4_edge": FeatureMeta(nth_edges_motif(4), {"me4"}),
}
if __name__ == "__main__":
from measure_tests.specific_feature_test import test_specific_feature
# Previous version contained a bug while counting twice sub-groups with double edges
# test_specific_feature(nth_edges_motif(3), is_max_connected=True)
test_specific_feature(nth_edges_motif(4), is_max_connected=True)
class FiedlerVectorCalculator(NodeFeatureCalculator):
def _calculate_dep(self, include: set):
# Working on every connected component by itself
self._features = dict(zip(self._gnx, alg_connectivity.fiedler_vector(self._gnx)))
def _calculate(self, include: set, is_regression=False):
self._features = {}
for graph in nx.connected_component_subgraphs(self._gnx):
if len(graph) < 2:
self._features.update(zip(graph.nodes(), [0.] * len(graph)))
else:
self._features.update(zip(graph.nodes(), map(float, alg_connectivity.fiedler_vector(graph))))
def is_relevant(self):
# Fiedler vector also works only on connected undirected graphs
# so if gnx is not connected we shall expect an exception: networkx.exception.NetworkXError
# return (not self._gnx.is_directed()) and (nx.is_connected(self._gnx.to_undirected()))
return not self._gnx.is_directed()
feature_entry = {
"fiedler_vector": FeatureMeta(FiedlerVectorCalculator, {"fv"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(FiedlerVectorCalculator, is_max_connected=True)
import networkx as nx
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class ClosenessCentralityCalculator(NodeFeatureCalculator):
def _calculate(self, include: set, is_regression=False):
self._features = nx.closeness_centrality(self._gnx)
def is_relevant(self):
return True
feature_entry = {
"closeness_centrality":
FeatureMeta(ClosenessCentralityCalculator, {"closeness"}),
}
if __name__ == "__main__":
from measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(ClosenessCentralityCalculator, is_max_connected=True)
import networkx as nx
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class AverageNeighborDegreeCalculator(NodeFeatureCalculator):
def is_relevant(self):
return True
def _calculate(self, include: set, is_regression=False):
self._features = nx.average_neighbor_degree(self._gnx)
feature_entry = {
"average_neighbor_degree":
FeatureMeta(AverageNeighborDegreeCalculator, {"avg_nd"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(AverageNeighborDegreeCalculator,
is_max_connected=True)
class EccentricityCalculator(NodeFeatureCalculator):
def _calculate(self, include: set, is_regression=False):
dists = {
src: neighbors
for src, neighbors in nx.all_pairs_shortest_path_length(self._gnx)
}
self._features = {
node: max(neighbors.values())
for node, neighbors in dists.items()
}
def _calculate_dep(self, include: set):
# Not using eccentricity to handle disconnected graphs. (If a graph has more than 1 connected components,
# the eccentricty will raise an exception)
self._features = {
node: nx.eccentricity(self._gnx, node)
for node in self._gnx
}
def is_relevant(self):
return True
feature_entry = {
"eccentricity": FeatureMeta(EccentricityCalculator, {"ecc"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(EccentricityCalculator, is_max_connected=True)
import networkx as nx
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class LoadCentralityCalculator(NodeFeatureCalculator):
def is_relevant(self):
return True
def _calculate(self, include: set, is_regression=False):
self._features = nx.load_centrality(self._gnx)
feature_entry = {
"load_centrality": FeatureMeta(LoadCentralityCalculator, {"load_c"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(LoadCentralityCalculator, is_max_connected=True)
min_degrees = min(graph.degree(), key=lambda x: x[1])[
1] # [deg for node, deg in graph.degree()])
isomap_mx = Isomap(
n_neighbors=min(min_degrees, MAX_DEGREE),
n_components=COMPONENT_SIZE).fit_transform(dissimilarities)
self._features.update(zip(nodes_order, isomap_mx))
@staticmethod
def _dissimilarity(graph, nodes_order):
m = nx.floyd_warshall_numpy(graph, nodelist=nodes_order)
return np.asarray(m)
feature_entry = {
"multi_dimensional_scaling":
FeatureMeta(MultiDimensionalScalingCalculator, {"mds"}),
}
def test_feature():
from graph_measures.loggers import PrintLogger
from graph_measures.measure_tests.test_graph import get_graph
gnx = get_graph()
feat = MultiDimensionalScalingCalculator(
gnx, logger=PrintLogger("Keren's Logger"))
res = feat.build()
print(res)
if __name__ == "__main__":
# from measure_tests.specific_feature_test import test_specific_feature
# Fast and numerically precise:
variance = np.average((values - average)**2, weights=weights)
return average, np.sqrt(variance)
def _calculate(self, include: set, is_regression=False):
for node in self._gnx:
# calculate BFS distances
distances = nx.single_source_shortest_path_length(self._gnx, node)
# distances.pop(node)
# if not distances:
# self._features[node] = [0., 0.]
# continue
node_dist = Counter(distances.values())
dists, weights = zip(*node_dist.items())
# This was in the previous version
# instead of the above commented fix
adjusted_dists = np.asarray([x + 1 for x in dists])
weights = np.asarray(weights)
self._features[node] = list(
self.weighted_avg_and_std(adjusted_dists, weights))
feature_entry = {
"bfs_moments": FeatureMeta(BfsMomentsCalculator, {"bfs"}),
}
if __name__ == "__main__":
from measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(BfsMomentsCalculator, is_max_connected=True)
from collections import Counter
import community
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class LouvainCalculator(NodeFeatureCalculator):
def is_relevant(self):
# relevant only for undirected graphs
return not self._gnx.is_directed()
def _calculate(self, include: set, is_regression=False):
partition = community.best_partition(self._gnx)
com_size_dict = Counter(partition.values())
self._features = {node: com_size_dict[partition[node]] for node in self._gnx}
feature_entry = {
"louvain": FeatureMeta(LouvainCalculator, {"lov"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(LouvainCalculator, is_max_connected=True)
self._features[element][r_type][x]
for x in self._gnx.graph["edge_labels"]
] for r_type in self._relation_types]).flatten()
# def _to_ndarray(self):
# mx = np.matrix([self._get_feature(node) for node in self._nodes()])
# return mx.astype(np.float32)
def nth_neighbor_calculator(order):
return partial(NthNeighborNodeEdgeHistogramCalculator, order)
feature_entry = {
"first_node_edge_histogram":
FeatureMeta(nth_neighbor_calculator(1), {"fnneh"}),
"second_node_edge_histogram":
FeatureMeta(nth_neighbor_calculator(2), {"snneh"}),
}
def sample_graph():
g = nx.DiGraph(edge_labels=[1, 2])
g.add_edges_from([
(1, 2, {
"label": 1
}),
(2, 6, {
"label": 2
}),
(4, 1, {
from graph_measures.features_infra.feature_calculators import NodeFeatureCalculator, FeatureMeta
class GeneralCalculator(NodeFeatureCalculator):
def is_relevant(self):
return True
def _calculate(self, include: set, is_regression=False):
if self._gnx.is_directed():
self._features = {node: (in_deg, out_deg) for
(node, out_deg), (_, in_deg) in zip(self._gnx.out_degree(), self._gnx.in_degree())}
else:
self._features = {node: deg for node, deg in self._gnx.degree()}
feature_entry = {
"general": FeatureMeta(GeneralCalculator, {"gen"}),
}
if __name__ == "__main__":
from graph_measures.measure_tests.specific_feature_test import test_specific_feature
test_specific_feature(GeneralCalculator, is_max_connected=True)