About Snowflake Notebooks¶
Snowflake Notebooks is a unified development interface in Snowsight that offers an interactive, cell-based programming environment for Python, SQL, and Markdown. In Snowflake Notebooks, you can leverage your Snowflake data to perform exploratory data analysis, develop machine learning models, and perform other data science and data engineering workflows, all within the same interface.
Explore and experiment with data already in Snowflake, or upload new data to Snowflake from local files, external cloud storage, or datasets from the Snowflake Marketplace.
Write SQL or Python code and quickly compare results with cell-by-cell development and execution.
Interactively visualize your data using embedded Streamlit visualizations and other libraries like Altair, Matplotlib, or seaborn.
Integrate with Git to collaborate with effective version control. See Sync Snowflake Notebooks with a Git repository.
Contextualize results and make notes about different results with Markdown cells and charts.
Run your notebook on a schedule to automate pipelines. See Schedule your Snowflake Notebook to run.
Make use of the role-based access control and other data governance functionality available in Snowflake to allow other users with the same role to view and collaborate on the notebook.
Notebook runtimes¶
Snowflake Notebooks offers two types of runtimes: warehouse runtime and container runtime. Notebooks rely on virtual warehouses and/or Snowpark Container Services compute pools to provide the compute resource. For both architectures, SQL and Snowpark queries are always pushed down to execute on the warehouse for optimized performance.
The warehouse runtime gives you the fastest way to start, with a familiar warehouse environment that is generally available. The container runtime provides a more flexible environment that can support many different types of workloads, including SQL analytics and data engineering. You can install additional Python packages if the container runtime doesn’t include what you need by default. Container runtime also comes in CPU and GPU versions that have many popular ML packages pre-installed, making them ideal for ML and deep learning workloads.
The following table shows supported features for each type of runtime. You can use this table to help decide which runtime is the right choice for your use case.
Supported Features |
Warehouse Runtime |
Container Runtime |
|---|---|---|
Compute |
Kernel runs on the notebook warehouse. |
Kernel runs on Compute Pool node. |
Environment |
Python 3.9 |
Python 3.10 |
Base image |
Streamlit + Snowpark |
Snowflake container runtime. CPU and GPU images are pre-installed with a set of Python libraries. |
Additional Python libraries |
Install using Snowflake Anaconda or from a Snowflake stage. |
Install using |
Editing support |
|
Same as warehouse |
Access |
Ownership required to access and edit notebooks. |
Same as warehouse |
Supported Notebook features (still in Preview) |
|
Same as warehouse |
Get started with Snowflake Notebooks¶
To start experimenting with Snowflake Notebooks, sign into Snowsight, set up your account to use notebooks,
and then select Notebooks from the Projects pane. A list of notebooks that you have access to in your account is displayed.
You can either create a new notebook from scratch or upload an existing .ipynb file.
The following table shows the topics to review if you’re new to Snowflake Notebooks:
Getting started guides |
|
|---|---|
|
Setting up Snowflake Notebooks Instructions for developers and admins before using Notebook. |
|
Create a new notebook from scratch or from an existing file. |
|
Develop and Run Code in Snowflake Notebook Create, edit, execute Python, SQL, Markdown cells. |
Developer guides¶
Guide |
Description |
|---|---|
Accessing and modifying the session context. |
|
Saving notebooks and results across sessions. |
|
Importing Python package from Anaconda channel. |
|
Visualize data with matplotlib, plotly, altair and develop a data app with Streamlit. |
|
Reference SQL cell output and Python variable values. |
|
Leverage keyboard shortcut to navigate and streamline editing experience. |
Leveling up your notebook workflows¶
Guide |
Description |
|---|---|
Version control your notebook for collaboration and development. |
|
Manage and work with files in your notebook environment. |
|
Schedule notebooks to run and execute code within Snowflake. |
|
Leverage other Snowflake capabilities within Snowflake Notebooks. |
|
Troubleshoot errors you may run into while using Snowflake Notebooks. |
Quickstarts¶
Getting Started with Your First Snowflake Notebook [Video] [Source]
Learn how to quickly get started with your first notebook project in less than 10 minutes.
Visual Data Stories with Snowflake Notebooks [Video] [Source]
Learn how you can create compelling data narratives using visualizations, Markdown, images, and interactive data apps all within your notebook, alongside your code and data.
Highlighted use cases¶
Check out highlighted use cases for data science, data engineering and ML/AI in Github.
|
Develop, manage, schedule and run scalable data pipelines with SQL and Snowpark. |
|
Explore, visualize, and analyze your data with Python and SQL. |
|
Feature Engineering, model training, and development with Cortex and Snowpark ML. |
Note
These quickstarts are only shown as examples. Following along with the example may require additional rights to third-party data, products, or services that are not owned or provided by Snowflake. Snowflake does not guarantee the accuracy of these examples.
Additional resources¶
For notebook demos, tutorials and examples, see collection of Snowflake Notebooks demos in GitHub.
To view tutorial videos, see the Snowflake Notebooks YouTube playlist.
To learn about SQL commands to create, execute and show notebooks, see Snowflake Notebooks API reference.
Looking for reference architectures, industry specific use-cases and solutions best practices using Notebooks? See Notebooks examples in the Snowflake Solution Center.