(no changes, no records) for similarity score

Hello!

If I try to compute the similarity score between different nodes which exists in my graph based on how many neighbors they share, for a subset of my nodes the result is (no changes, no records), even thought they do have some common neighbors.

I am following the same logic as presented here: Node Similarity - Neo4j Graph Data Science.

What could be the problem for those nodes with common neighbors, but with "(no changes, no records)" result for similarity score? Thank you very much in advance!

Hi @anaoprea - can you share the Cypher statements(s) you are trying? I'm assuming you are using write mode, something like `CALL gds.nodeSimilarity.write(...)?

Hello and thank you very much for your help! Here is the query:

CALL gds.nodeSimilarity.stream('FINAL')
YIELD node1, node2, similarity
where (gds.util.asNode(node1).name= "Solyc09g00792021" and
gds.util.asNode(node2).name= "AT2G370401")
RETURN gds.util.asNode(node1).name AS Gene1, gds.util.asNode(node2).name AS Gene2, similarity

which returns (no changes, no records)...

If you remove this predicate from the query do you get any results?

where (gds.util.asNode(node1).name= "Solyc09g00792021" and
gds.util.asNode(node2).name= "AT2G370401")

Hello again!

Yes, I get results. To frame the problem again:

I am interested for similarity between nodes (A and B) and (C and D). They all exist and the two pairs have nodes in common. A and C are the same type. B and D are the same type. If I compute similarity for (A and B) I will get a result. If I compute similarity for (C and D) I will get (no changes, no records), even thought I should have a number just like in the case of (A and B)...