Neo4J Bloom加载Python导入数据后不显示节点属性及分类问题

问题根因 #

你编写的导入代码没有为节点设置标签（Label），Neo4j Bloom的透视图功能依赖节点标签做分类识别，无标签的节点只会被识别为通用的node分类，无法自动匹配对应属性做展示。而Neo4j浏览器不强制依赖标签，所以可以正常显示节点属性。

解决步骤 #

方案1：修改导入代码重新导入（适合还没正式投入使用的新库） #

修改Python代码中创建节点的Cypher语句，为不同类型的节点加上对应标签，示例修改后代码如下：

from neo4j import GraphDatabase

driver = GraphDatabase.driver("connection", auth=("neo4j", "password"))

def add_data(tx):
    tx.run("LOAD CSV WITH HEADERS FROM 'file:///C:/Users/Damian/PycharmProjects/NeoJ/DataMap.csv' AS Map \
    MERGE (source:DataNode {node_name: Map.source}) \
    MERGE (destination:DataNode {node_name: Map.destination}) \
    CREATE (source)-[:FEEDS_INTO]->(destination)")

def add_other(tx):
    tx.run("LOAD CSV WITH HEADERS FROM 'file:///C:/Users/Damian/PycharmProjects/NeoJ/Data.csv' AS Data \
        Match (node:DataNode {node_name: Data.dane}) \
        MERGE (system:System {system_name: Data.system}) \
        MERGE (scope:Scope {scope_name: Data.scope}) \
        MERGE (process:Process {process_name: Data.process}) \
        MERGE (owner:Owner {owner_name: Data.owner}) \
        MERGE (node)-[:UNDER_SYSTEM]->(system) \
        MERGE (system)-[:UNDER_SCOPE]->(scope) \
        MERGE (node)-[:HAS_PROCESS]->(process) \
        MERGE (owner)-[:IS_OWNER_OF]->(node) ")

def add_data_properties(tx):
    tx.run("LOAD CSV WITH HEADERS FROM 'file:///C:/Users/Damian/PycharmProjects/NeoJ/Data.csv' AS Data \
    MATCH (n:DataNode {node_name: Data.dane }) \
    SET n.system = Data.system \
    SET n.scope = Data.scope \
    SET n.process = Data.process \
    SET n.owner = Data.owner")

with driver.session() as session:
    session.write_transaction(add_data)
    session.write_transaction(add_other)
    session.write_transaction(add_data_properties)
driver.close()

方案2：直接给现有节点补标签（不需要重新导数据） #

在Neo4j浏览器中依次执行以下Cypher语句，为已有节点补充标签：

// 给数据节点加DataNode标签
MATCH (n) WHERE EXISTS(n.node_name) SET n:DataNode;
// 给系统节点加System标签
MATCH (n) WHERE EXISTS(n.system_name) SET n:System;
// 给域节点加Scope标签
MATCH (n) WHERE EXISTS(n.scope_name) SET n:Scope;
// 给流程节点加Process标签
MATCH (n) WHERE EXISTS(n.process_name) SET n:Process;
// 给负责人节点加Owner标签
MATCH (n) WHERE EXISTS(n.owner_name) SET n:Owner;

Bloom端配置 #

完成标签补充后，重新生成Bloom透视图，此时透视图会自动识别出DataNode、System等多个节点分类，你可以在透视图的分类设置中，为每个分类勾选需要在节点上展示的属性即可正常显示。

内容的提问来源于stack exchange，提问作者neekitit