How to use notebooks in Azure Data Studio
This article describes how to launch the Notebook experience in Azure Data Studio and how to start authoring your own notebooks. It also shows how to write Notebooks using different kernels.
Connect to SQL Server
You can connect to the Microsoft SQL Server connection type in Azure Data Studio. In Azure Data Studio, you can also press F1, and click New Connection and connect to your SQL Server.
There are multiple ways to launch a new notebook.
Go to the File Menu in Azure Data Studio and then click on New Notebook.
Right click on the SQL Server connection and then launch New Notebook.
Open the command palette (Ctrl+Shift+P)) and then type in New Notebook. A new file named
Supported kernels and attach to context
The Notebook Installation in Azure Data Studio natively supports SQL Kernel. If you're a SQL developer and want to use Notebooks, then this would be your chosen Kernel.
The SQL Kernel can also be used to connect to PostgreSQL server instances. If you're a PostgreSQL developer and would like to connect to your PostgreSQL Server, then download the PostgreSQL extension in the Azure Data Studio extension Marketplace.
In the code cells within the Notebook, similar to our query editor, we support modern SQL coding experience that makes your everyday tasks easier with built-in features such as a rich SQL editor, IntelliSense, and built-in code snippets. Code snippets allow you to generate the proper SQL syntax to create databases, tables, views, stored procedures, etc., and to update existing database objects. Use code snippets to quickly create copies of your database for development or testing purposes and to generate and execute scripts.
Click Run to execute each cell.
SQL Kernel to connect to SQL Server instance
SQL Kernel to connect to PostgreSQL Server instance
Configure Python for Notebooks
When you select any of the other kernels apart from SQL from the kernel dropdown, this prompts you to Configure Python for Notebooks. The Notebook dependencies get installed in a specified location but you can decide whether to set the installation location. This installation can take some time and it's recommended to not close the application until the installation is complete. Once the installation is complete, you can start writing code in the supported language.
Once the installation succeeds, you can find a notification in the Task History along with the location of the Jupyter backend server running in the Output Terminal.
|SQL Kernel||Write SQL Code targeted at your relational database.|
|PySpark3 and PySpark Kernel||Write Python code using Spark compute from the cluster.|
|Spark Kernel||Write Scala and R code using Spark compute from the cluster.|
|Python Kernel||Write Python code for local development.|
Attach to provides the context for the Kernel to attach. If you're using SQL Kernel, then you can
Attach to any of your SQL Server instances.
If you're using Python3 Kernel the
Attach to is
localhost. You can use this kernel for your local Python development.
When you're connected to SQL Server 2019 big data cluster, the default
Attach to is that end point of the cluster that lets you submit Python, Scala, and R code using the Spark compute of the cluster.
Code Cells and Markdown Cells
Add a new code cell by clicking the +Code command in the toolbar.
Add a new text cell by clicking the +Text command in the toolbar.
The cell changes to edit mode and now type markdown and you can see the preview at the same time
Clicking outside the text cell shows the markdown text.
Trusted and Non Trusted
Notebooks open in Azure Data Studio are default Trusted.
If you open a Notebook from some other source, it opens in Non Trusted mode and then you can make it Trusted.
You can save the Notebook by Ctrl+S or clicking the File Save, File Save As... and File Save All commands from the File menu and File: Save commands entered in the command palette.
PySpark Kernel and in the cell type in the following code.
The Spark Application is started and returns the following output:
Spark kernel | Scala language
Spark|Scala Kernel and in the cell type in the following code.
You can also view the "Cell Options" when you click on the options icon below –
Spark kernel | R language
Choose the Spark | R in the dropdown for the kernels. In the cell, type or paste in the code. Click Run to see the following output.
Local Python kernel
Choose the local Python Kernel and in the cell type in -
One of the things we optimized for local Python development was to include the ability to install packages which customers would need for their scenarios. By default, we include the common packages like
numpy etc., but if you're expecting a package that isn't included then write the following code in the notebook cell:
When you run this command,
Module not found is returned. If your package exists, then you don't get the error.
If it returns a
Module not Found error, then click on Manage Packages to launch the wizard experience.
In this wizard, you're able to see the Installed packages. You can search through the list and the associated version of each of these packages. If you need to uninstall any of these packages, then you can click on one of the packages and then click on the Uninstall selected packages option.
You can also click on Add new packages to Search for a particular package, choose the related version and click install. By default, we select the latest version of the searched package.
After the package is installed, you can go in the Notebook cell and type in following command:
If you need to uninstall any of these packages, then you can click on one or multiple packages and then click on the Uninstall selected packages option.
To learn how to work with an existing notebook, see How to manage notebooks in Azure Data Studio.