My question is very similar to this one: How to create unique nodes and relationships by csv file imported in neo4j? I have a textfile with around 2.5 million lines that has two columns, each one being node ids:
1234 345
1234 568
345 984
... ...
Each line represents a relationship (so 2.5 million relationships): first_column nodeid-> FOLLOWS -> second_column nodeid. There are around 80,000 unique nodes in this file.
Based on the link above, I did:
USING PERIODIC COMMIT 1000
LOAD CSV FROM 'file:///home/user_name/Desktop/bigfile.csv' AS line FIELDTERMINATOR ' '
MERGE (n:Userid { id: toInt(line[0]) })
WITH line, n
MERGE (m:Userid { id: toInt(line[1]) })
WITH m,n
MERGE (n)-[:FOLLOWS]->(m)
I am assuming this code
My main question is I am wondering how to make this process faster. This is being done on Ubuntu, and I changed the values from 512 to 2048 MB for memory in the conf/neo4j-wrapper.conf file. (maximum I can increase on my Virtual Machine)
Should I try doing the Import tool? Based on example on this website, neo4j.com/developer/guide-import-csv/ under "Super Fast Batch Importer For Huge Datasets",
./bin/neo4j-import --into mydatabase.db --id-type INTEGER \
--nodes allnodes.csv \
--delimiter " " \
--relationships:FOLLOWS bigfile.csv
And to do this, I need to reformat files so that: allnodes.csv shows
userID:ID(Userid)
1234
5678
...
And bigfile.csv shows
:START_ID(Userid) :END_ID(Userid)
1234 345
1234 568
345 984
*Two columns delimited by space*
And when I run this import, I get this error:
Input error: Expected '--nodes' to have at least 1 valid item, but had 0 []
Caused by:Expected '--nodes' to have at least 1 valid item, but had 0 []
java.lang.IllegalArgumentException: Expected '--nodes' to have at least 1 valid item, but had 0 []
How do I fix this error? And for the csv files, do I put them in same folder where I run this command (neo4j folder)?
Your command line probably has the wrong paths for your two CSV files.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With