Development environment
Taking the time to properly set up your development environment can save time and headaches in the long run. Your development environment will usually consist of a mix of local software and remote services. Consider taking the following practices into account for future work.
Local Development Environment#
General Tips#
Docker#
Using Docker comes with lot of advantages, one of the most prominent ones being the ability to run programs in an environment that is always the same (and thus, no more 'it works on my laptop' issues). Consider using Docker to run existing third party software on your laptop, to share your program with others, or to run your application on a remote server.
Conda#
Consider using Conda environments to make managing libraries for different programming languages easier. Be wary of the difference between pip and Conda if you do.
Terminal basics#
The terminal is the interface with which you can execute text based commands. We recommend getting a basic familiarity to at least be able to navigate the folder structure and run shell commands. This is not as difficult as it may seem at first (windows cmd will not run some of these commands, see the section below to set up a similar terminal experience).
List files and directories in the current directory:
Change directory:
Get the current directory you are in:
The easiest way to open and edit a file in the terminal is to use nano:
For purists
vim and emacs are also available.
Windows#
Updates#
For security and usability reasons, it is recommended to continuously update your Windows OS to the latest version.
Install Windows Subsystem for Linux#
Installing Windows Subsystem for Linux will allow you to run Ubuntu Linux in your Windows environment. This is the recommended approach for any running .sh scripts you may find in other repositories. Furthermore, WSL is recommended as a best practice in favour of the environment native cmd and powershell. This provides better reusability as the same scripts can be used on Linux and MacOs. Follow the official instructions to install WSL 2 on your machine.
Install with the command line
A simpler solution is available only if you signed up for Windows 10 insider to access development builds of Windows:
Different environments
Most applications installed using the WSL terminal will only be available in WSL.
Microsoft terminal#
Microsoft terminal provides an improved user experience compared to the default terminals such as Ubuntu Shell, PowerShell, CMD. See the instructions to install the new Microsoft Terminal
Installing packages#
WSL package manager#
You can use the apt package manager to install packages on WSL.
When you want to install a package just google "install MY_APPLICATION ubuntu", you should be proposed a few easy way to install your application/package in a few commands.
We recommend the installation of zsh and oh-my-zsh, it will take a minute and upgrade your Ubuntu terminal. Providing a much more user-friendly way to navigate the terminal.
Winget#
Install winget, the official Windows package manager. Still in beta at the moment. So the list of available packages is limited.
Chocolatey#
For packages not available in winget we recommend to use Chocolatey, the most popular tool to install packages on Windows.
Remote Environments#
Data Science Research Infrastructure#
The Data Science Research Infrastructure (DSRI) at Maastricht University provides CPU and GPUs nodes for Data Science computing. This service should be considered for all research that requires a large amount of computational resources, as well as anything that benefits from being run remotely and/or continuously. At the moment it is less suitable for public web services, as it requires a VPN connection to the UM network to gain access. Any program or script can easily be run on DSRI by deploying a Docker container.
info
See the DSRI documentation
Collaborative environments#
For free computational resources, collaborative scripting and sharing your work, consider using cloud environments such as Google Collab, Kaggle or a combination of Jupyter Notebook and Binder.
Free power
Kaggle is free during competitions.
Store data#
Sensible research data#
If you have sensible research data: use DataHub
Research data#
If you have data to publish related to a research project: use SURF