This tutorial will take you through installing Docker and ContainDS on a compatible machine, then launching a simple virtual environment running Jupyter.
Since ContainDS uses Docker for its underlying virtual machines (‘containers’) you need to install and run Docker on your computer, in addition to ContainDS.
Installing Docker Desktop
All you need to have running is the ‘Docker daemon CE’ (Community Edition). The package to install is Docker Desktop. You can locate the download on the Docker website (may require free sign up to their ‘Docker Hub’ service), or follow the direct links here:
- Docker Desktop for Windows (requires Windows 10 64-bit: Pro, Enterprise, or Education – more details here)
- Docker Desktop for Mac (requires macOS 10.12 or newer, hardware from 2010 or newer – more details here)
If your computer does not satisfy the requirements for Docker Desktop above, you will need to install Docker Toolbox instead. See our frequently asked questions for details.
Download the installer for ContainDS itself:
- ContainDS for Windows
- ContainDS for Mac
Select the appropriate installer from our Download page.
Your operating system may require you to confirm you wish to trust Ideonate Ltd.
Double click the ContainDS icon or locate it within your Applications folder.
If Docker isn’t running you may see this screen:
Running Docker Desktop
Locate and start the Docker Desktop application.
To check the status of Docker look for the whale icon in the system tray or notification area of your operating system:
Docker Desktop on Windows
Docker Desktop on Mac
Once Docker is fully running, you should be able to click ‘Retry Setup’ in ContainDS and the full setup should proceed correctly.
Launch a Jupyter Container
Once Docker is running correctly, you should see the main ‘New Container’ screen in ContainDS:
From the list of recommended Jupyter images look for ‘datascience-notebook’. Click ‘SELECT’ next to that image.
On the next screen you can choose a different name for the new container (i.e. for the environment/workspace pair) if you like.
You do need to input the workspace folder – where on your computer’s drive you would like your notebook files to be stored. This can be an existing folder which already contains notebook files or other data, perhaps cloned from a git repo. Below, we’ve entered the full path for ‘workspace’ inside the user’s home folder, which will be created automatically if it doesn’t already exist.
Click ‘CREATE’. You will see a progress animation while the image is downloaded from Docker Hub. When the container is ready, you will see the Jupyter console logs:
Click on the ‘WEB’ button to launch Jupyter Lab in your default web browser:
You can now use Jupyter Lab to create or view ipynb notebooks as normal. To install packages into the environment, just use Terminal within Jupyter Lab itself:
For full details on possibilities for starting new containers see New Container Options.
And don’t forget to sign up to our newsletter to find out about new features and guides!