Import CSV data into Neo4j graph database inside a Docker container

Prerequisites:

• a working Docker Neo4j container
• a CSV file of denormalized data sourced from a relational database
• a graph database model derived from your relational schema data
• a prepared Cypher script that will push CSV data into Neo4j
• a clean graph.db database

---

What can I use to import data into Neo4j?

There are a number of tools that we can use to import external data into a Neo4j graph:

Neo4j Browser - it will run LOAD CSV statements but only one at a time

neo4j-shell - is a command line utility that comes pre-installed with Neo4j and will run multi-statement Cypher scripts to run against a graph database. Each statement must be terminated with a semicolon(;) and separated with an empty line

neo4j-import - is a command line utility that comes pre-installed with Neo4j and is designed for bulk loading massive datasets that exceed 10 million records. You can also use this tool to test far smaller datasets. However, note that your CSV files must follow a very specific format[¹]

LazyWebCypher - an online web app that will run multi-statement Cypher scripts even against your own local Neo4j instance[²]

cycli - Cypher command-line interface[³]

How to import data into Neo4j using neo4j-shell

1. Confirm that the CSV file, soil_survey_sample.csv, is in place

   sudo head -3 neo4j/import/soil_survey_sample.csv 
   Hort_Client,Contractor,Region,Locality,Soil_Service,Solution,Soil_Issue,Date_Reported,Date_Actioned,DaysToAction
   159,1091,Northbury,3656,54593,5397,Erosion,2007-05-07,2008-02-18,287
   159,1091,Northbury,1516,22644,5397,Erosion,2007-05-07,2008-03-18,316
   

2. Confirm that the Cypher file, soil_survey_import_to_neo4j_in_docker.cql, is in place

   sudo head -2 neo4j/import/soil_survey_import_to_neo4j_in_docker.cql 
   CREATE INDEX ON :Hort_Client(client);
   CREATE INDEX ON :Hort_Client(name);
   

3. In a terminal different to the one that Neo4j service is running in, enter the following command

   sudo docker exec -ti $(sudo docker ps --format '{{.Names}}') bin/neo4j-shell -file import/soil_survey_import_to_neo4j_in_docker.cql
   

   Sample output:

   • +-------------------+
     | No data returned. |
     +-------------------+
     Indexes added: 1
     66 ms
     +-------------------+
     | No data returned. |
     +-------------------+
     Indexes added: 1
     11 ms
     +-------------------+
     | No data returned. |
     +-------------------+
     Nodes created: 21
     Properties set: 42
     Labels added: 21
     723 ms
     +-------------------+
     | No data returned. |
     +-------------------+
     Indexes added: 1
     19 ms
     +-------------------+
     | No data returned. |
     +-------------------+
     Indexes added: 1
     15 ms
     +-------------------+
     | No data returned. |
     +-------------------+
     Nodes created: 2670
     Properties set: 5340
     Labels added: 2670
     980 ms
     +-------------------+
     | No data returned. |
     +-------------------+
     Relationships created: 2796
     5457 ms
     ..
     ..
     ...
     

4. Get a list of node labels

   sudo docker exec -ti $(sudo docker ps --format '{{.Names}}') bin/neo4j-shell -c "CALL db.labels();"
   

   Sample output:

   • +----------------+
     | label          |
     +----------------+
     | "Hort_Client"  |
     | "Soil_Service" |
     | "Solution"     |
     | "Soil_Issue"   |
     | "Soil_Report"  |
     | "Contractor"   |
     | "Region"       |
     | "Locality"     |
     +----------------+
     8 rows
     15 ms
     

5. Get a list of relationships

   sudo docker exec -ti $(sudo docker ps --format '{{.Names}}') bin/neo4j-shell -c "CALL db.relationshipTypes();"
   

   Sample output:

   • +------------------+
     | relationshipType |
     +------------------+
     | "REQUESTS"       |
     | "RECOMMENDS"     |
     | "HAS"            |
     | "INVESTIGATES"   |
     | "CORRECTS"       |
     | "DISCUSSES"      |
     | "ISSUES"         |
     | "SENT_TO"        |
     | "ACTIONS"        |
     | "WORKS_AT"       |
     | "OPERATES_IN"    |
     | "PART_OF"       |
     +------------------+
     12 rows
     10 ms
     

6. Generate a meta-graph in Neo4j Browser

   CALL db.schema();
   

   Soil Survey meta-graph:

   • 

---

You have successfully populated Neo4j database using neo4j-shell utility and confirmed existence of new nodes and relationships

---

Back to top of page

---