Creating a relationship from csv

Neo4j Graph Platform Cypher
1

I have created nodes and relationships from .csv files:

LOAD CSV WITH HEADERS FROM "file:///Ancestry_match.csv" AS matches 
CREATE (m1:DNAmatch {
matchid: matches.matchGuid, 
matchname: matches.matchTestDisplayName, 
matchadminname: matches.matchTestAdminDisplayName, 
matchcM: matches.sharedCentimorgans})


LOAD CSV WITH HEADERS FROM "file:///Ancestry_AncestorList.csv" AS thrulines 
CREATE (m2:thru {
pathfrommetoancestor: thrulines.KinshipPathToSampleId, 
ancestorsname: thrulines.AncestorDisplayName})


LOAD CSV WITH HEADERS FROM "file:///AncestryGreatGrandParents.csv" AS ggParent 
CREATE (m3:ggp {
pathfrommetomyggp: ggParent.ggpID, 
myggP: ggParent.ggpName})


CREATE INDEX match_index IF NOT EXISTS FOR (m1:DNAmatch) ON (m1.matchid)


MATCH (dnamatch:DNAmatch)
SET dnamatch.matchcM = toFloat(dnamatch.matchcM)


LOAD CSV WITH HEADERS FROM "file:///AncestryAncestorCouple.csv" AS thruline 
MATCH (a:ggp {pathfrommetomyggp: thruline.ggpID}) 
MATCH (b:thru {pathfrommetoancestor: thruline.KinshipPathToSampleId}) 
MERGE (a)-[:lineof]->(b)


LOAD CSV WITH HEADERS FROM "file:///Ancestry_Ancestor.csv" AS matchtoAncestor 
MATCH (a1:DNAmatch {matchid: matchtoAncestor.matchGuid}) 
MATCH (b1:thru {pathfrommetoancestor: matchtoAncestor.KinshipPathToSampleId}) 
MERGE (a1)-[:pathtoancestor]->(b1)

All of the nodes load, as well as, the index and float conversion. The first relationship (MERGE (a)-[:lineof]->(b)) works and when returning those nodes show properly. The second relationship responds with (no changes, no records). If I return a1 or b1 I get the proper node info.

I can't see where the difference is between the two relationships or why the second one won't work.

Thanks for any help
Dean

2

The code does look correct. Does the relationship already exist?

3

No, I have wiped the db, reloaded just the two node sets, then the relationship and get this. When I return the a1 or b1 (with limit like 5) I get the nodes. Then wipe the db again and it doesn't list eliminating relationships.

4

I take that back. Using MATCH (X) RETURN X putting a1, b1 or any of those it returns the same nodes. I apparently am using a misconception of how that works.

5

This is resolved. Look at data files. One file has ID's all caps the other all lower case.

Thank you for looking at this Gary.

1 Like