SQLMesh is a data transformation framework that brings the benefits of DevOps to data teams. It enables data scientists, analysts, and engineers to efficiently run and deploy data transformations written in SQL or Python.
If you're new to SQLMesh, it may be helpful to walk through the quickstart guide before working with the SQLMesh example project.
For more information, check out the website and documentation.
This repository contains standalone example SQLMesh projects for you to tinker with, test, and break. The examples vary in size and complexity, demonstrating different SQLMesh features and options.
Each set of example projects contains a Jupyter notebook (.ipynb file) that walks through the project's contents and describes notable features. The notebook file is not necessary to run/use the example project and can be deleted. You can view a rendered notebook in Github if you do not want to run it locally (e.g., the sushi project overview notebook).
SQLMesh projects require data to transform, and each project uses a local DuckDB database to store its data. Each example contains a database with a small amount of data so the projects will run out of the box.
The projects each contain a Python file with helper functions that add data to the database, which is useful for simulating real-world scenarios where new data are ingested over time. Each helper file defines a command line interface, so the functions may be called from the CLI or by importing and calling them in Python (as the overview notebook files do).
Running the examples requires three steps:
- Clone/download this git repository
- Install SQLMesh
- Execute the projects
Running the example projects requires cloning this Git repository and installing Python libraries.
To clone the repository, first ensure that Git is installed on your computer. In your command line interface (CLI), navigate to the folder where you want the sqlmesh-examples directory to be placed. Issue the git clone https://github.com/TobikoData/sqlmesh-examples.git command to create a copy of the repo on your machine:
Next, navigate to the sqlmesh-examples directory in your CLI. We recommend creating a Python virtual environment with the command python3 -m venv .venv, then activating it with the command source .venv/bin/activate:
Running the examples requires a few different Python libraries:
- SQLMesh (including web UI dependencies if you want to use the graphical interface) [required to do anything]
- Jupyter notebooks [required to run overview notebooks]
Install only the necessary libraries:
pip install sqlmesh
Or install all the libraries:
pip install "sqlmesh[web]" notebook
Navigate to the directory for one of the example projects, and you are ready to go!
After navigating to a project directory, you can open the project in the SQLMesh user interface with the sqlmesh ui command, work from the CLI with commands like sqlmesh plan, or open a notebook by issuing the jupyter notebook command (or using an IDE like VSCode's native notebook functionality).
Have fun!