CODEWITHSUNDEEP
neo4jcypher
Priyanka Basu
Priyanka Basu

Reputation: 407

Nodes having highest and lowest connections across unique node properties in Neo4j

I have two types of nodes in a graph DB. One type is Testplan and the other is Tag.

I want to find which Tag nodes that have most connections across unique Testplan.TP_MNEM (this is a property of Testplan node).

To explain it in a bit more detail:

Let’s say the graph has 5 Testplan nodes, 3 with TP_MNEM as “AAA”, 1 with TP_MNEM as “BBB” and the other with TP_MNEM as “CCC”.

I want the Tag nodes that are connected to most types of TP_MNEM. If a Tag Node 1 has 3 connections with TP_MNEM “AAA” and no connection with any other TP_MNEM then Tag Node 1 should get lower priority over Tag Node 2 which has 2 connections but each with unique TP_MNEM (“AAA” and “CCC”, for example).

Similarly I want the Tag nodes having highest degree but are only connected to only one type of TP_MNEM. Here Tag Node 1 will get precedence over Tag Node 2.

The code to upload a CSV is shared below:

USING PERIODIC COMMIT
LOAD CSV WITH HEADERS FROM "file:///C:/SRA_clean.csv" AS row

MERGE(testplan:Testplan{
TP_ID:row.TP_ID,
DB_NAME:row.DATABASE_NAME,
TP_MNEM:row.TP_MNEMONIC,
TEST_TYPE:row.TEST_TYPE,
TEST_PLAN_NAME:row.TEST_PLAN_NAME,
CREATION_DATE:row.TS_CREATION_DATE,
MODIFIED_DATE:row.TS_MODIFIED_DATE,
SQL_SERVE_UPD_DATE:row.UPDATE_DT_TM_SQL_SERV,
MYSQL_UPD_DATE:row.UPDATE_DT_TM_MY_SQL
})
FOREACH (tagName IN split(row.Keywords,",")|
MERGE (tag:Tag{name:tagName})
MERGE(testplan)-[:TAGGED]->(tag))

Any help would be great.

Upvotes: 1

Views: 48

Answers (1)

Tom Geudens
Tom Geudens

Reputation: 2666

I may have simplified things (and/or misunderstood your questions) ... but doesn't the following query fulfill your first question ?

MATCH (tg:Tag)<-[:TAGGED]-(tp:Testplan)
RETURN tg.tagName, count(DISTINCT tp.TP_MNEM) AS count ORDER by count DESC;

And the other question ... well ... just replace DESC with ASC ...

Hope this helps.

Regards, Tom

Upvotes: 0

Related Questions