Return all nodes reachable from a list of nodes and their `maximal` paths

[lavanya_kannan2]

Hi,

I am trying to see if there is an APOC procedure that will return all nodes reachable from a list of nodes and return only all the maximal paths from the input list. Currently, I am able to use the apoc.path.expand procedure to get all the paths from the input list of nodes. From this, I am post-processing and extracting only those maximal paths (paths that contain both the nodes and relationships of other paths and those that end at the same node). I am wondering if there can be a procedure that remembers this maximal requirement while traversing the graph and returns only the maximal paths. Thanks.

[lavanya_kannan2]

Update: The method I am looking for is similar to the method 2 in Find all reachable nodes from every node present in a given set - GeeksforGeeks, but for directed graphs.

Lavanya

[Bennu]

Hi @lavanya.kannan2 !

How performant should it be? And more important... How confident do you feel about extending APOC? This looks like an specific requirement that could be easily attacked by creating your own procedure on the DB.

Take a look on

Neo4j Graph Database Platform

Let’s Write a Stored Procedure in Neo4j – Part I

I am frequently asked by Neo4j customers how to improve query performance on large graphs. Neo4j offers two main ways of querying a graph – through the Cypher query language and via a Java API.

Bennu

[lavanya_kannan2]

Hi,

The new procedure should perform better than the approach of collecting all paths and then deciding. For small number of nodes, it would not be a great performance difference. However, for large amounts of input nodes that may be reachable from each other, I feel there would definitely be an improvement. I will look into writing my own procedure, when time and expertise permits. Thanks for the link.

Lavanya