Do you have an index or constraint on :nodes(id) ?
If not create one
Don't do this
> db.getNodeById(db.findNode(Labels.nodes, "id",
> "9149892904013846978").getId()))
but
> db.findNode(Labels.nodes, "id", "9149892904013846978")
Find the nodes once upfront and assign them to variables
and not during every step of the travesal.
Don't use a list for lookup but a Set
You can use Uniqueness.NODE_PATH to assure that a node only appears once in a
path
Only ONE Uniqueness can be used.
As you continually check for nodes on each branch in the evaluator, you only
have to test the endNode
return includeNodesSet.contains(path.endNode()) ? return INCLUDE_AND_CONTINUE :
EXCLUDE_AND_CONTINUE;
What does createPath do?
your procedure doesn't return anything?
You don't use the start-point.
Try to leave off the sorting and see if it helps.
Run the procedure multiple times to see what the warmed up time is.
Add a counter to your evaluator and output that counter so you see how many
Paths it touches.
If possible you should limit it to certain rel-types or labels.
> Am 31.01.2018 um 14:37 schrieb Miosota <popho...@gmail.com>:
>
> Yes, i didn't use RestGraphDB in my plugin. I used it in my test project for
> connection from Intellij idea project to neo4j database. Database consists of
> about 50K nodes and 100K relationships.
>
>
> I'm running neo4j 3.3.1. For plugin creation use neo4j-kernel 3.3.1 .
>
> For test i created a procedure below. It finds paths from one point to
> another including at least one point from a set of 3 points. Traverser spend
> too much time to find this paths. 17-20s.
> And i really need bidirectional. And the graph later will be much more bigger
> (millions elements).
>
> May be i need to change smth in neo4j properties? I dont know.
> My RAM 8G.
> dbms.memory.heap.initial_size=1000m
> dbms.memory.heap.max_size=2000m
> dbms.memory.pagecache.size=5000m
>
> @Procedure(value = "mytraverser.pathIncludingAtLeastOneNode")
> @Description("createIncludingAtLeastOneNode")
> public void createIncludingAtLeastOneNode(@Name("startPoint") String
> startPoint, @Name("endPoint") String endPoint) {
>
> try (Transaction tx = db.beginTx()) {
> TraversalDescription td = db.traversalDescription().depthFirst()
> .uniqueness(Uniqueness.NODE_PATH)
> .uniqueness(Uniqueness.RELATIONSHIP_PATH)
> .relationships(RelationshipTypes.rel, Direction.BOTH)
> .evaluator(Evaluators.includingDepths(0, 10));
>
> Traverser tr = td
>
> .evaluator(Evaluators.includeWhereEndNodeIs(db.getNodeById(db.findNode(Labels.nodes,
> "id", "9149892904013846978").getId())))
> .evaluator(new PathEvaluator.Adapter() {
> @Override
> public Evaluation evaluate(Path path, BranchState
> branchState) {
> if (path.length()==0) return
> Evaluation.INCLUDE_AND_CONTINUE;
> else {
> ArrayList<Node> listNode = new ArrayList<>();
> for (Node n:path.nodes()){
> listNode.add(n);
> }
> if
> (listNode.contains(db.getNodeById(db.findNode(Labels.nodes, "id",
> "9149892904013846972").getId())) ||
>
> listNode.contains(db.getNodeById(db.findNode(Labels.nodes, "id",
> "9149892904013846980").getId())) ||
>
> listNode.contains(db.getNodeById(db.findNode(Labels.nodes, "id",
> "9149892904013846974").getId()))) {
> return Evaluation.INCLUDE_AND_CONTINUE;
> } else {
> return Evaluation.EXCLUDE_AND_CONTINUE;
> }
> }
> }
> })
> .sort(new PathSorting("weight"))
> .traverse(db.getNodeById(db.findNode(Labels.nodes, "id",
> "9149892904413851091").getId()));
>
> createPath(tr, "IncludingAtLeastOneNode");
> tx.success();
> }
> }
>
> среда, 31 января 2018 г., 14:59:09 UTC+3 пользователь Michael Hunger написал:
> Can you share your full code somewhere?
>
> RestGraphDB should not be used.
>
> Which version are you runnign on?
>
> Your plugin doesn't do anyting you have to actually use the traverser to
> extract the nodes or paths.
>
>
>
> Do your really need 10 levels of bidirectional (BOTH!) relationships? That is
> potentially billions of paths.
> You can add a counter in the relationship-expander to see how many rels are
> hit.
>
> Can't you just use shortestpath?
>
>
>
> > Am 31.01.2018 um 10:45 schrieb Svetlana Usacheva <poph...@gmail.com>:
> >
> > Hello! I have a problem with my plugin. Firstly, i tried work with my
> > database of 150K elements from my app using RestGraphDatabase for
> > connection. Implementation takes about 3 s. Then i built jar and put it in
> > folder "plugins". So now i can use my procedures from plugin, but they work
> > much slower (about 17 s). Can i speed up my plugin and how i can do it?
> >
> > Procedures looks like that:
> >
> > import org.neo4j.procedure.Context;
> > import org.neo4j.procedure.Description;
> > import org.neo4j.procedure.Name;
> > import org.neo4j.procedure.Procedure;
> >
> > public class MyTraverser{
> >
> > @Context
> > public GraphDatabaseService db;
> >
> > @Procedure(value = "path")
> > @Description("createPath")
> > public void createPath(@Name("startPoint") String startPoint,
> > @Name("endPoint") String endPoint) {
> >
> > try (Transaction tx = db.beginTx()) {
> > TraversalDescription td =
> > db.traversalDescription().depthFirst()
> > .uniqueness(Uniqueness.NODE_PATH)
> > .relationships(RelationshipTypes.rel, Direction.BOTH)
> > .evaluator(Evaluators.includingDepths(0, 10));
> >
> > Traverser tr = td
> > .evaluator(Evaluators.includeWhereEndNodeIs(endPoint)
> > .traverse(startPoint);
> >
> > tx.success();
> > }
> > }
> > }
> >
> > --
> > You received this message because you are subscribed to the Google Groups
> > "Neo4j" group.
> > To unsubscribe from this group and stop receiving emails from it, send an
> > email to neo4j+un...@googlegroups.com.
> > For more options, visit https://groups.google.com/d/optout.
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "Neo4j" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to neo4j+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups
"Neo4j" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to neo4j+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
