Workbench: Your Gateway to Exploring Descartes Labs Platform APIs with JupyterLab

Getting Started with Workbench

Let's Begin

Embark on your Workbench journey by logging in to https://app.descarteslabs.com/workbench, you'll be greeted by the hub home interface. Here, you can initiate, manage, and connect to your JupyterLab server. To launch your server, simply click the "Start My Server" button. You'll then be prompted to select the type of compute instance you require: "Model Exploration/Development" or "Model Training." Choose the environment that aligns with your project's needs. Remember, you can easily switch between server types without losing your progress, as your home directory will persist across instance changes.

Getting Set Up

To ensure a smooth and secure connection to the Descartes Labs Platform APIs through the Python client library, you'll need to perform a one-time initialization of your environment. This process involves setting up an authentication token for future API calls.

Navigate to the "example-notebooks/" directory in your JupyterLab environment's filesystem browser. Inside this directory, you'll find a notebook named "01 Logging In.ipynb." Open the notebook and follow the provided instructions to add your credentials to your JupyterLab home directory. If you prefer, you can also use the command line. Simply open a terminal within JupyterLab and execute:

descarteslabs auth login

The terminal will guide you through the necessary steps.

Exploring Workbench's Features

Pre-loaded Examples

Upon starting your server, you'll discover a collection of pre-loaded examples in the "example-notebooks" folder within your home directory. These examples serve as excellent starting points for acquainting yourself with the capabilities of the Descartes Labs Platform APIs and visualization tools. Keep in mind that these examples are refreshed each time your server starts, potentially overwriting any contents in the "example-notebooks" directory. For this reason, it's advisable to refrain from saving your work in that directory.

Managing Server Lifecycle

The JupyterLab server you initiate will be monitored for inactivity. If it remains inactive for a period, it may be automatically shut down. Should this happen, don't worry—it's easy to restart your server by following the same process you used to start it initially. Notably, you won't need to re-initialize your environment.

If you ever need to manually shut down your server, navigate to the hub home at https://app.descarteslabs.com/workbench/hub/home and click the "Stop My Server" button. Within a few seconds, the server will shut down. At this point, you can either log out or begin a new server session, depending on your next steps.

Conclusion

Descartes Labs Workbench empowers you to seamlessly engage with the Descartes Labs Platform APIs and perform advanced geospatial analysis within a user-friendly JupyterLab environment. Start exploring the endless possibilities today! If you encounter any questions or challenges, remember that this knowledge base is here to assist you on your journey.