Importing Relational Data into Neo4j ›Importing Data
Import Northwind Nodes

Importing Relational Data into Neo4j

Foundation

Designing the Graph Model

Importing Data

Validating and Querying

Import Northwind Nodes

Your challenge is to use the Neo4j Data Importer to upload Northwind CSV files, model your data, and create nodes.

To complete this challenge:

  1. Upload the Northwind CSV files to Data Importer

  2. Create node labels for Customer, Order, Product, Category, Supplier, Employee, and Shipper

  3. Map CSV columns to node properties and set unique identifiers

  4. Run the import to create nodes in your database

Neo4j Data Importer main interface

Choosing Your Import Environment

Access the Data Importer via AuraDB:

  1. Log in to Neo4j Aura at console.neo4j.io/graphacademy

  2. Open your AuraDB instance

  3. Click Import in the left sidebar

The Data Importer opens directly connected to your Aura database.

Need to clear an existing model?

If you have an existing data importer model, you can clear it by selecting the Clear all option from the menu …​.

The clear all option in the data importer menu '…​'

Creating Nodes

Create nodes with Data Importer:

  1. Add a data source (CSV file)

  2. Create a node label

  3. Map properties from the CSV file to the node

  4. Set a unique identifier

  5. Run the import

Step 1: Add a Data Source

Data Importer can import data from CSV (Comma Separated Values) and TSV (Tab Separated Values) files.

  1. Click New data source in the Files panel

  2. Select CSV as the data source type

    Add new data source

  3. Drag and drop your CSV file or click to browse

    Drop CSV file

  4. Once uploaded, the file contents will be displayed

    CSV file contents displayed

For this course, you can use the Northwind CSV files from: https://github.com/neo4j-graph-examples/northwind/tree/main/import

Step 2: Create a Node Label

Start with a blank canvas:

Blank data model canvas

Create new nodes:

  1. Click Add node label on the canvas

  2. Enter the label name such as Customer, Product, or Order

  3. In the Definition panel, select the corresponding CSV file

  4. Click Map from table to map CSV columns to node properties

  5. Select the columns you want to import and click Confirm

Node with mapped properties

Editing properties

You can edit property names and types by clicking the pencil icon next to each property. This allows you to:

  • Rename columns to follow Neo4j naming conventions, for example customer_id to customerID

  • Change data types: String, Integer, Float, Boolean, Date

  • Remove columns you do not need

Type mismatch in Data Importer

Type mismatch in Data Importer

If Data Importer cannot convert a source value to the specified data type (e.g. text in a column you set as Integer), the import continues successfully but that node will not have that property. Check your data types and verify node counts if you suspect type mismatches.

Step 3: Set a Unique Identifier

Every node type needs a unique identifier to:

  • Prevent duplicate nodes during import

  • Enable efficient lookups

  • Create relationships between nodes

Click the key icon next to the ID property to set it as the unique identifier.

The key icon next to the ID property

Data Importer creates a Name for each property and assigns a Type based on the data in the file. The Column refers to the field in the CSV file.

When you set a unique identifier, Data Importer automatically creates a unique constraint and index for that property. This enables MERGE semantics on re-import and faster lookups.

Constraints and Indexes in Data Importer

You can view and manage constraints and indexes from the Data Importer UI:

  1. Select a node in your model

  2. Switch to the Constraints & Indexes tab in the Definition panel

  3. The unique ID property creates a constraint and index automatically

  4. To add an extra index for query performance (e.g. on companyName for Customer), click the + button under Indexes, select the property, and run the import

Constraints & Indexes tab in Data Importer

Adding indexes improves query performance for filters and lookups. For example, an index on orderDate speeds up date-range queries on Order nodes.

Step 4: Run the Import

Once your model is complete:

  1. Click Run import

    The run import button

  2. Wait for the import to complete. You will see a summary of the results:

    Import results summary

  3. Click Close to return to Data Importer

Step 5: Save Your Model

After a successful import:

  1. Give your data model a name such as "Northwind Model"

  2. Click Save

Named data model

Reusing your model

Saved models can be reused for future imports. You can also download your model using the …​ menu > Download model (with data) to back it up or share with colleagues.

Verify the Import

Verify your data in the Query tool:

cypher
// Count nodes by label
MATCH (n)
RETURN labels(n)[0] AS label, COUNT(*) AS count
ORDER BY count DESC
cypher
// View sample data
MATCH (c:Customer) RETURN c LIMIT 10
Query results showing imported nodes

Common Import Mistakes to Avoid

Common mistakes to avoid:

Mistake 1: Not setting a unique identifier

Problem: Importing nodes without a unique ID property.

Consequence: * Cannot use MERGE to update existing nodes * May create duplicate nodes on re-import * Difficult to create relationships later

Solution: Always set a unique identifier for each node type before importing.

Mistake 2: Importing foreign keys as properties

Problem: Keeping foreign key columns as node properties.

Example: Creating a Customer node with customerID AND keeping order_id as a property.

Reality: Foreign keys should become relationships, not properties. The order_id on a customer does not make sense - instead, create a relationship from Customer to Order.

Mistake 3: Wrong data types

Problem: Importing all data as strings when other types are appropriate.

Examples: * Dates stored as strings instead of Date type * Numbers stored as strings instead of Integer/Float * Booleans stored as "true"/"false" strings

Solution: Check and set appropriate data types in Data Importer or convert during Cypher import.

Mistake 4: Creating nodes before constraints

Problem: Importing nodes first, then trying to add constraints.

Consequence: If duplicate data exists, constraint creation will fail.

Solution: Always create constraints BEFORE importing data:

cypher
CREATE CONSTRAINT customer_id IF NOT EXISTS
FOR (c:Customer) REQUIRE c.customerID IS UNIQUE

Backup your work

Data Importer saves changes automatically, but you can download your import model and data by selecting the Download model (with data) option from the menu …​.

The download model (with data) option in the data importer menu '…​'

You can restore your model using the Open model (with data) option and selecting the downloaded file.

Need to download a working solution?

This Data Importer model, person-import.zip, contains a working solution for this exercise.

Download the model and open it using the Open model (with data) button in the …​ menu.

The Open model (with data) button highlighted in the menu.

Note that this will replace your existing model.

Check Your Understanding

Creating Nodes with Data Importer

What must you do to create nodes using Data Importer? (Select all that apply)

  • ✓ Upload a source file

  • ✓ Create a node label

  • ✓ Set a unique identifier for each node

  • ❏ Update at least one property name

Hint

You can modify a node’s property names, although it is not mandatory.

Solution

To create nodes using Data Importer, you must:

  • Upload a source file

  • Create a node label

  • Set a unique identifier for each node.

You do not have to update any property names.

Summary

In this challenge, you imported Northwind nodes using the Data Importer. In the next lesson, you will create relationships between them.

Chatbot

How can I help you today?