Edit

Share via

Quickstart: Create your first graph in Microsoft Fabric

Note

This feature is currently in public preview. This preview is provided without a service-level agreement, and isn't recommended for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see Supplemental Terms of Use for Microsoft Azure Previews.

In this quickstart, you create a basic graph with two nodes and one edge, and then run a query. This process takes about 15 minutes.

For a comprehensive tutorial that covers the full Adventure Works data model and more advanced scenarios, see the Graph tutorial.

Prerequisites

Before you start this quickstart, verify that:

  1. Graph is available in your region.

  2. Graph is enabled in your Fabric tenant.

    Enable graph in your Fabric tenant.

  3. You're a member of a Fabric workspace or have permission to create items in the workspace. For more information, see Workspaces in Microsoft Fabric.

    Important

    Access management of the graph is restricted to the workspace that hosts it. Users outside of the workspace can't access the graph. Users within the workspace who have access to the underlying data in the lakehouse can model and query the graph.

Load sample data

To create your graph, first load sample data into a lakehouse in your Fabric workspace.

Download the sample data

  1. Go to the Fabric Graph GQL example datasets on GitHub.

  2. Select the adventureworks_docs_sample.zip file and download it to your local machine.

    Tip

    To download a file from GitHub, select the file, and then select the Download raw file icon.

  3. Extract the downloaded adventureworks_docs_sample.zip file to a folder on your local machine.

    Tip

    In File Explorer, right-click the zip file and select Extract All, then choose a destination folder.

Create a lakehouse

  1. In Microsoft Fabric, select your workspace.

  2. Select + New item.

  3. Select Store data > Lakehouse.

  4. Enter a name (for example, "AdventureWorksLakehouse"), clear the Lakehouse schemas option, and select Create.

    Important

    Graph doesn't currently support lakehouses with lakehouse schema (preview) enabled.

Load the data into tables

For this quickstart, you only need two tables: adventureworks_customers and adventureworks_orders. Upload the full sample data folder, and then load just these two tables.

  1. In your lakehouse, hover over Files, select the ellipsis (...), and then select Upload > Upload folder.
  2. Browse to the extracted folder and upload it. This action uploads all the sample data files to your lakehouse.
  3. Expand Files and the uploaded folder to see the subfolders. For this quickstart, you only need to load two of them as tables.
  4. Hover over the adventureworks_customers subfolder, select the ellipsis (...), and choose Load to Tables > New table.
  5. Set the file type to Parquet and select Load.
  6. Repeat steps 4-5 for the adventureworks_orders subfolder.

Create a graph model

  1. In your Microsoft Fabric workspace, select + New item.

  2. Select Analyze and train data > Graph model (preview).

    Screenshot showing the new item menu with the option to select Graph model (preview).

    Tip

    Alternatively, enter "graph" in the search box and press Enter to search for graph items.

  3. Enter a name (for example, "MyFirstGraph") and select Create.

After creating the graph model, you're taken to the graph view where you can see the default mode is set to Model with Save, Get data, Add node, and Add edge buttons at the top.

Create a graph

To create a graph in Microsoft Fabric, follow these steps in graph view:

  1. Select Get data.

  2. From the OneLake catalog, select your lakehouse with the Adventure Works data.

  3. Select Connect.

  4. Select the adventureworks_customers and adventureworks_orders tables, and then select Load.

    Screenshot showing the get data dialog with selected tables.

Add two nodes

  1. Select Add node.

  2. Configure the first node:

    • Label: Customer
    • Mapping table: adventureworks_customers
    • ID of mapping column: CustomerID_K

    Screenshot showing the add node to graph dialog.

  3. Select Confirm.

  4. Select Add node again.

  5. Configure the second node:

    • Label: Order
    • Mapping table: adventureworks_orders
    • ID of mapping column: SalesOrderDetailID_K

  6. Select Confirm.

Add one edge

  1. Select Add edge.

  2. Configure the edge:

    • Label: purchases
    • Mapping table: adventureworks_orders
    • Source node: Customer
    • Source mapping column: CustomerID_FK
    • Target node: Order
    • Target mapping column: SalesOrderDetailID_K

    Screenshot showing the add edge dialog.

  3. Select Confirm.

  4. Select Save to load the graph.

Query your graph

Run a GQL query to find the top five customers by order count.

  1. Select Query mode.

  2. Select Code editor from the top menu.

  3. Enter the following GQL query in the input box:

    MATCH (c:Customer)-[:purchases]->(o:`Order`)
RETURN c.fullName AS customer_name, count(o) AS num_orders
GROUP BY customer_name
ORDER BY num_orders DESC
LIMIT 5

  4. Select Run query to see the top five customers by order count.

The following image shows the GQL query and its results:

Screenshot showing the result of running a GQL query.

Congratulations! You created your first graph in Microsoft Fabric and ran a query against it.

Next steps

  • Last updated on