如何在Neo4j Browser中将查询到的路径保存为新图？

Great question! Unfortunately, Neo4j doesn’t have a built-in MERGE path command that automatically replicates your matched paths into a new subgraph, but there are a few simple, practical ways to get this done—depending on whether you want to keep the subgraph in the same database or export it to a completely new one.

Option 1: Mark the Subgraph in the Same Database #

If you just want to flag the nodes and relationships from your matched paths so you can easily reference them later (instead of creating duplicates), add a custom label or attribute to identify them:

MATCH path = (n:A {title:"WATER"})-[r*2]-(m:B)
// Add a label to all nodes in the path
FOREACH (node IN nodes(path) | SET node:WaterSubgraph)
// Add a property to all relationships in the path
FOREACH (rel IN relationships(path) | SET rel.part_of_water_subgraph = true)

Now you can quickly query this subgraph anytime with:

MATCH path = (n:WaterSubgraph)-[r:part_of_water_subgraph*]-(m:WaterSubgraph)
RETURN path

Option 2: Export to a New Database (As a Standalone Graph) #

If you need a separate, independent copy of the subgraph, use Neo4j’s APOC library (you’ll need to have it installed and enabled first) to export the paths, then import them into a new database.

Step 1: Export the Matched Paths #

Export the subgraph to a GraphML file (easy to import later):

MATCH path = (n:A {title:"WATER"})-[r*2]-(m:B)
CALL apoc.export.graphml.data(nodes(path), relationships(path), "water_subgraph.graphml", {})
YIELD file, nodes, relationships
RETURN file, nodes, relationships

You can also export to a Cypher script instead, which lets you run direct create statements:

MATCH path = (n:A {title:"WATER"})-[r*2]-(m:B)
CALL apoc.export.cypher.data(nodes(path), relationships(path), "water_subgraph.cypher", {format: "cypher-shell"})
YIELD file, nodes, relationships
RETURN file, nodes, relationships

Step 2: Import into a New Database #

1. Create a new database in your Neo4j instance:

   CALL db.createDatabase('water_subgraph_db')
   

2. Switch to the new database, then run the import command:

   CALL apoc.import.graphml("water_subgraph.graphml", {})
   

   Or if you exported a Cypher script, just run the contents of water_subgraph.cypher in the new database.

Quick Alternative: Duplicate Nodes/Relationships in the Same Database #

If you want to create exact duplicates of the nodes and relationships (not just mark them), use this query to copy all properties and labels/relationship types:

MATCH path = (n:A {title:"WATER"})-[r*2]-(m:B)
WITH nodes(path) AS originalNodes, relationships(path) AS originalRels
// Create duplicate nodes and store their original ID -> new node mapping
UNWIND originalNodes AS origNode
CREATE (newNode)
SET newNode = origNode, newNode:DuplicateSubgraph
WITH collect({origId: id(origNode), newNode: newNode}) AS nodeMap, originalRels AS origRels
// Create duplicate relationships using the node mapping
UNWIND origRels AS origRel
MATCH (startNew) WHERE startNew = [nm IN nodeMap WHERE nm.origId = id(startNode(origRel))][0].newNode
MATCH (endNew) WHERE endNew = [nm IN nodeMap WHERE nm.origId = id(endNode(origRel))][0].newNode
CREATE (startNew)-[newRel:type(origRel)]->(endNew)
SET newRel = origRel
RETURN count(newRel) AS duplicated_relationships

This will create a separate copy of your subgraph with the :DuplicateSubgraph label for easy identification.

内容的提问来源于stack exchange，提问作者Asya