Is it possible to get only one node "gary" by using header name?

[sreeraj_sj]

[glilienfield]

Sorry, I don’t understand the question. Will you please provide more detail? 

[sreeraj_sj]

the csv file contain duplicate nodes and null values . a and b are persons and there relation is c. I'm trying to get a o/p without duplication and null values and along with that i need a single node 'Gary' which comes under the header B.   here is the o/p -> 

 now i need one more node  which is "Gary".

[sreeraj_sj]

i used this query,
i need one more node "gary " with its property "e"
load csv with headers from "file:///newtask.csv" as row

WITH row WHERE NOT row.E IS null

WITH row WHERE NOT row.A IS null

MERGE(p1:person{firstName:row.A,age:row.E})

MERGE(p2:person{firstName:row.B})

WITH p1, p2, row

CALL apoc.create.relationship(p1, row.C,{number:row.D}, p2) YIELD rel

RETURN rel

[glilienfield]

You filtered that row out, since A is null. What do you want to do with B, since A is null and the relationship can’t be created between A and B 

[ameyasoft]

Try this:
load csv with headers from "file:///newtask.csv" as row
WITH row WHERE NOT row.E IS null

CALL 

{
	WITH row
	WITH row WHERE NOT row.A IS null
	MERGE(p1:Person{firstName:row.A,age:row.E})
	return p1, "yes" as flg

UNION

	WITH row
	WITH row WHERE row.A IS null and row.B is not null
	MERGE(p1:Person{firstName:row.B,age:row.E})
	return p1, "no" as flg
}

WITH row, p1, flg 
MERGE(p2:Person{firstName:row.B})

WITH row, p2, p1, flg where flg = "yes"
CALL apoc.create.relationship(p1, row.C,{number:row.D}, p2) YIELD rel
RETURN rel

[sreeraj_sj]

thank you

return p1, "yes" as flg

can you please explain me this part, it is confusing