Closed ugurdogrusoz closed 3 years ago
In this case, it is executed undirected. So we go from seed to some other node and we come back to the seed node. For our algorithm, they are valid cases currently. If it was directed, I think we don't see those.
Here actually the nodes are highlighted with 3 different colors. The seeds are blue, the targets (regulators) are green. The things in the middle are yellow. Nodes with labels "Jennifer Lopez" and "Ellen Barkin" are also highlighted as common targets (regulators). I think this is valid in the undirected case.
How can you go to Lopez from Freeman using 2 edges?
It seems like I forgot to update the JAR file or maybe restart the database to activate the newest java implementations. I updated the JAR files and restart Neo4j databases under "\~/visuall/neo4j-server-hydra" and "\~/visuall/neo4j-server-base" Now it looks OK.
Some notes
I see unexpected results here.
Load the attached file for CTR query. (loads Fred Astaire and Easter Parade) fred_astaire-easter_parade.txt
Then execute CTR with length limit 2 and undirected I see the below image.
Why "Top Hat" comes as a common target/regulator? Do we not show the path from "Top Hat" to "Easter Parade"?
-- Edit: Since it is "undirected" "Top Hat" connects to "Easter Parade" with the path (Top Hat)-(Fred Astaire)-(Easter Parade). So I think the results are right.
My original example works fine now. Example provided by @canbax is OK as well since the seed nodes have a direct edge between them, both nodes' immediate neighbors are common targets of these nodes in 2 steps.
Why do we have nodes/edges outside the red area highlighted here? Besides, some actresses are marked as if they are in the result (outside this area).