Del via


Tutorial: Run Python on a cluster and as a job using the Databricks extension for Visual Studio Code

This tutorial walks you through setting up the Databricks extension for Visual Studio Code, and then running Python on an Azure Databricks cluster and as an Azure Databricks job in your remote workspace. See What is the Databricks extension for Visual Studio Code?.

Requirements

This tutorial requires that:

  • You have installed the Databricks extension for Visual Studio Code. See Install the Databricks extension for Visual Studio Code.
  • You have a remote Azure Databricks cluster to use. Make a note of the cluster’s name. To view your available clusters, in your Azure Databricks workspace sidebar, click Compute. See Compute.

Step 1: Create a new Databricks project

In this step, you create a new Databricks project and configure the connection with your remote Azure Databricks workspace.

  1. Launch Visual Studio Code, then click File > Open Folder and open some empty folder on your local development machine.
  2. On the sidebar, click the Databricks logo icon. This opens the Databricks extension.
  3. In the Configuration view, click Migrate to a Databricks Project.
  4. The Command Palette to configure your Databricks workspace opens. For Databricks Host, enter or select your per-workspace URL, for example https://adb-1234567890123456.7.azuredatabricks.net.
  5. Select an authentication profile for the project. See Authentication setup for the Databricks extension for Visual Studio Code.

Step 2: Add cluster information to the Databricks extension and start the cluster

  1. With the Configuration view already open, click Select a cluster or click the gear (Configure cluster) icon.

    Configure cluster

  2. In the Command Palette, select the name of the cluster that you created previously.

  3. Click the play icon (Start Cluster) if it is not already started.

Step 3: Create and run Python code

  1. Create a local Python code file: on the sidebar, click the folder (Explorer) icon.

  2. On the main menu, click File > New File. Name the file demo.py and save it to the project’s root.

  3. Add the following code to the file and then save it. This code creates and displays the contents of a basic PySpark DataFrame:

    from pyspark.sql import SparkSession
    from pyspark.sql.types import *
    
    spark = SparkSession.builder.getOrCreate()
    
    schema = StructType([
       StructField('CustomerID', IntegerType(), False),
       StructField('FirstName',  StringType(),  False),
       StructField('LastName',   StringType(),  False)
    ])
    
    data = [
       [ 1000, 'Mathijs', 'Oosterhout-Rijntjes' ],
       [ 1001, 'Joost',   'van Brunswijk' ],
       [ 1002, 'Stan',    'Bokenkamp' ]
    ]
    
    customers = spark.createDataFrame(data, schema)
    customers.show()
    
    # Output:
    #
    # +----------+---------+-------------------+
    # |CustomerID|FirstName|           LastName|
    # +----------+---------+-------------------+
    # |      1000|  Mathijs|Oosterhout-Rijntjes|
    # |      1001|    Joost|      van Brunswijk|
    # |      1002|     Stan|          Bokenkamp|
    # +----------+---------+-------------------+
    
  4. Click the Run on Databricks icon next to the list of editor tabs, and then click Upload and Run File. The output appears in the Debug Console view.

    Upload and run file from icon

    Alternatively, in the Explorer view, right-click the demo.py file, and then click Run on Databricks > Upload and Run File.

    Upload and run file from context menu

Step 4: Run the code as a job

To run demo.py as a job, click the Run on Databricks icon next to the list of editor tabs, and then click Run File as Workflow. The output appears in a separate editor tab next to the demo.py file editor.

Run file as workflow from icon

Alternatively, right-click the demo.py file in the Explorer panel, then select Run on Databricks > Run File as Workflow.

Run file as workflow from context menu

Next steps

Now that you have successfully used the Databricks extension for Visual Studio Code to upload a local Python file and run it remotely, you can also: