How to create a relation between two existing nodes only when the relation has some value

vijji90 · November 13, 2019, 12:14pm

How to create a relation between two existing nodes only when the relation has some value. Eg., I have node labels Juice, Sour, sweet, spicy, salty. I will connect Sweetlime juice node to Sour node with relation[:HAS {value:30%}] and to sweet node with relation[:HAS {value:60%}] and to salty node with relation[:HAS {value:10%}] . Similarly I want to connect banana juice to Sweet node with relation[:HAS {value:100%}].

I have used below query
LOAD CSV WITH HEADERS FROM "file:///juice.csv" as line
MERGE (j:Juice {name:line.Juice})
WITH j,line
MATCH (a:Taste_Type {name:"sour"}),(b:Taste_Type {name:"salt"}),(c:Taste_Type {name:"sweet"}),(d:Taste_Type {name:"spicy"})
MERGE (j)-[r:HAS]->(a) SET r.value=line.Sour
MERGE (j)-[s:HAS]->(b) SET s.value=line.Salt
MERGE (j)-[t:HAS]->(c) SET t.value=line.Sweet
MERGE (j)-[u:HAS]->(d) SET u.value=line.Spicy

When I use the above query it is creating relationship between juice node and each and every taste_type node though the relation has no value in csv. I dont want sweet lime to connect to spicy node and banana juice connect to salty,sour,spicy. Could you help me how can I write the query to achieve this.

Thanks in Advance.

lju · November 14, 2019, 7:03am

Hi!

Could you please share with me what the first few lines of your CSV look like, including header?

Thanks!

vijji90 · November 14, 2019, 8:07am

juice

sura-nsgindia · November 14, 2019, 9:01am

Give it a try:

LOAD CSV WITH HEADERS FROM "file:///juice.csv" as line
MERGE (j:Juice {name:line.Juice})
WITH j,line
MATCH (a:Taste_Type {name:"sour"}),(b:Taste_Type {name:"salt"}),(c:Taste_Type {name:"sweet"}),(d:Taste_Type {name:"spicy"})
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sour) && line.Sour IS NOT NULL THEN [1] ELSE [] END |
    MERGE (j)-[r:HAS]->(a) SET r.value=line.Sour
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Salt) && line.Salt IS NOT NULL THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(b) SET r.value=line.Salt
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sweet) && line.Sweet IS NOT NULL THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(c) SET r.value=line.Sweet
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Spicy) && line.Spicy IS NOT NULL THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(d) SET r.value=line.Spicy
);

vijji90 · November 19, 2019, 5:13am

Thank you for the response. I tried above query. But it didnt work. It created nodes but not any of the relations.

sura-nsgindia · November 19, 2019, 5:18am

change this condition for each foreach to appropriate one and try again

vijji90 · November 19, 2019, 5:40am

I have used the below query when I tried.

LOAD CSV WITH HEADERS FROM "file:///juice.csv" as line
MERGE (j:Juice {name:line.Juice})
WITH j,line
MATCH (a:Taste_Type {name:"sour"}),(b:Taste_Type {name:"salt"}),(c:Taste_Type {name:"sweet"}),(d:Taste_Type {name:"spicy"})
FOREACH(ignoreMe IN CASE WHEN line.Sour IS NOT NULL THEN [1] ELSE END |
MERGE (j)-[r:HAS]->(a) SET r.value=line.Sour
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Salt) and line.Salt IS NOT NULL THEN [1] ELSE END |
MERGE (j)-[s:HAS]->(b) SET s.value=line.Salt
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sweet) and line.Sweet IS NOT NULL THEN [1] ELSE END |
MERGE (j)-[t:HAS]->(c) SET t.value=line.Sweet
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Spicy) and line.Spicy IS NOT NULL THEN [1] ELSE END |
MERGE (j)-[u:HAS]->(d) SET u.value=line.Spicy
)

Is it correct?

sura-nsgindia · November 19, 2019, 7:22am

Check this one

LOAD CSV WITH HEADERS FROM "file:///juice.csv" as line
MERGE (j:Juice {name:line.Juice})
WITH j,line
MATCH (a:Taste_Type {name:"sour"}),(b:Taste_Type {name:"salt"}),(c:Taste_Type {name:"sweet"}),(d:Taste_Type {name:"spicy"})
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sour)  THEN [1] ELSE [] END |
    MERGE (j)-[r:HAS]->(a) SET r.value=line.Sour
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Salt)  THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(b) SET r.value=line.Salt
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sweet)  THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(c) SET r.value=line.Sweet
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Spicy) THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(d) SET r.value=line.Spicy
);

vijji90 · November 19, 2019, 9:23am

It didn't work too. I tried by removing EXISTS() and having only NOT NULL condition in FOREACH also. But it didn't work too.

vijji90 · November 25, 2019, 1:04pm

Did you find any solution for this?

sura-nsgindia · November 27, 2019, 5:04am

Can share the CSV file

vijji90 · November 27, 2019, 6:00am

I shared the csv in above conversation. I cannot upload csv file as it allows only png, jpeg formats. Hence I attached screenshot of csv in above conversation.

sura-nsgindia · November 28, 2019, 5:37am

hi,
I have tested the code

Here what I have used

I have created four nodes

CREATE (:Taste_Type {name:"sour"});
CREATE (:Taste_Type {name:"salt"});
CREATE (:Taste_Type {name:"sweet"});
CREATE (:Taste_Type {name:"spicy"});

link to csv: juice.csv

Now I have created the relations and juice using the following query

LOAD CSV WITH HEADERS FROM "file:///juice.csv" as line
MERGE (j:Juice {name:line.Juice})
WITH j,line
MATCH (a:Taste_Type {name:"sour"}),(b:Taste_Type {name:"salt"}),(c:Taste_Type {name:"sweet"}),(d:Taste_Type {name:"spicy"})
WITH j, line, a, b, c, d
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sour)  THEN [1] ELSE [] END |
    MERGE (j)-[r:HAS]->(a) SET r.value=line.Sour
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Salt)  THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(b) SET r.value=line.Salt
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sweet)  THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(c) SET r.value=line.Sweet
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Spicy) THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(d) SET r.value=line.Spicy
);

Taste_Type nodes: Taste_Type nodes

relations: relations

Here a better version of the relation query:

USING PERIODIC COMMIT
LOAD CSV WITH HEADERS FROM "file:///juice.csv" as line
MERGE (j:Juice {name:line.Juice})
WITH j,line
MATCH (a:Taste_Type {name:"sour"})
MATCH (b:Taste_Type {name:"salt"})
MATCH (c:Taste_Type {name:"sweet"})
MATCH (d:Taste_Type {name:"spicy"})
WITH j, line, a, b, c, d
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sour)  THEN [1] ELSE [] END |
    MERGE (j)-[r:HAS]->(a) SET r.value=line.Sour
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Salt)  THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(b) SET r.value=line.Salt
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Sweet)  THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(c) SET r.value=line.Sweet
)
FOREACH(ignoreMe IN CASE WHEN EXISTS(line.Spicy) THEN [1] ELSE [] END | 
    MERGE (j)-[r:HAS]->(d) SET r.value=line.Spicy
)

Let me know if this is what you needed.