【发布时间】:2015-01-02 06:52:19
【问题描述】:
如何从networkx图中提取随机节点?我有一个 networkx 图形式的地图,我必须从中提取 5 个随机节点并获取与每个节点及其边缘相关联的数据。我想我可以用“np.random.choice”做到这一点,但我仍然无法得到任何结果。
【问题讨论】:
-
请发布您的代码。
标签: python random nodes networkx
如何从networkx图中提取随机节点?我有一个 networkx 图形式的地图,我必须从中提取 5 个随机节点并获取与每个节点及其边缘相关联的数据。我想我可以用“np.random.choice”做到这一点,但我仍然无法得到任何结果。
【问题讨论】:
标签: python random nodes networkx
import networkx as nx
from random import choice
g = nx.Graph()
g.add_edge(1,2)
g.add_edge(1,3)
g.add_edge(1,4)
g.add_edge(1,5)
g.add_edge(5,6)
random_node = choice(g.nodes())
【讨论】:
choice(list(g.nodes))。
请检查sample 方法,而不是使用choice。
from random import sample
import os
import networkx as nx
# load a graph from local
working_path = r'XXX'
graph = nx.read_gexf(os.path.join(working_path, 'XXX.gexf'))
# random sample 3 nodes from the graph
random_nodes = sample(list(graph.nodes()), 3)
如果您有一个具有不同节点类型的图形g,您可以使用以下函数来计算具有特定类型的节点数:
def count_nodes_with_type(graph, attribute_name, query):
"""
Count the number of nodes with specific type
:param graph: a networkx graph whose nodes have several different types
:param attribute_name: the attribute name used to access different type of nodes
:param query: the search query
:return: number of nodes satisfying the query
"""
node_attribute_dict = nx.get_node_attributes(graph, attribute_name)
filtered_nodes = {key: value for (key, value) in node_attribute_dict.items() if value == query}
return len(filtered_nodes)
您可以使用相同的逻辑来使用nx.get_edge_attributes 方法计算特定类型边的数量。
【讨论】:
对于 NetworkX 的最新版本(我认为是>= 2.5),您可以直接在节点视图上使用random.sample() 来获取节点的标签/索引或标签和数据的样本节点。
import networkx as nx
import random as rd
# Generate the example Karate club graph provided in NetworkX
g = nx.karate_club_graph()
print(g.nodes) # Output: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33]
# Get a random sample (without replacement) of the node labels/indexes:
sample = rd.sample(g.nodes, 3)
print(sample) # Output: [22, 18, 6]
# Get a random sample (without replacement) of the node labels and data:
sample = rd.sample(g.nodes.items(), 3)
print(sample) # Output: [(24, {'club': 'Officer'}), (27, {'club': 'Officer'}), (31, {'club': 'Officer'})]
在稍旧的版本中(从2.0 到2.5 之前),您需要在使用random.sample 之前将节点视图转换为列表。
注意:如果您安装了最新的 NetworkX 软件包,则不必担心这一点。
# Get a random sample (without replacement) of the node labels/indexes
# in older version of NetworkX.
sample = rd.sample(list(g.nodes), 3)
【讨论】: