MERGE duplicate relationships

[VilladsClaes]

This CYPHER

MATCH(a) WHERE ID(a) =1 MATCH(b) WHERE ID(b) = 2 CREATE (n)-[r]->(l)

of course results in duplicate relationships when run twice

which CYPHER should run to merge the duplicate relationships into one, without affecting the nodes?

[dana_canzano]

I'm not sure this works as your cypher is

MATCH(a) WHERE ID(a) =1 MATCH(b) WHERE ID(b) = 2 CREATE (n)-[r]->(l)

and your create references a node aliased as n and one aliased as l but I do not see where n or l defined.

Are you simply trying to create 1 relationship between node a and node b and if the above is run multiple times it should always result in only 1 relationship between a and b? Also, your relationship has no type. Is that what you want?

Maybe something similar to

MATCH(a) WHERE ID(a) =1 MATCH(b) WHERE ID(b) = 2 CREATE (a)-[:MY_RELATIONSHIP_TYPE]->(b)

[clem · Accepted solution]

I think this will help, but alas, the page doesn't have any examples, so you'll have to experiment:

Neo4j Graph Database Platform

apoc.refactor.mergeRelationships - APOC Documentation

This section contains reference documentation for the apoc.refactor.mergeRelationships procedure.

This doc also doesn't explain the configure argument, but the config arguments are described with MergeNodes:

https://neo4j-contrib.github.io/neo4j-apoc-procedures/3.5/graph-refactoring/merge-nodes/