Setting up an Environment

The ease of getting started on a project is demonstrated every time UCL runs its Machine Learning MSc courses on SherlockML. Within 5 minutes of starting the session, 200 participants have signed up and have a fully provisioned Data Science environment ready with Jupyter Notebooks and the entire scientific coding stack.

More experienced data scientists will enjoy being able to use the command line to run code and scripts, all of which can be hosted in SherlockML. Users can install packages and interact with the workspace (which also has a full GUI).

SherlockML also provides custom server environments. Custom environments let you define sets of Python packages to be installed with conda and pip, and system packages to be installed with APT. When creating a new server, choose an environment to apply and the server will start with all the specified packages already installed.

Even better: any team members in the project can also apply the environment to their server, so you’ll never again need to email a list of arcane shell commands.

For advanced users, in addition to Python and system packages, custom environments also let you specify Bash scripts to be run when the environment is applied, allowing arbitrary customisation.

Power your Data Science with SherlockML