How to set up a working data science environment in OpenSuse

OpenSuse is a great linux operating system that is easy to install, and works out of the box. It is a stable, functional linux system. It has a graphical installer and installation is a breeze. You can install it alongside other operating systems, but I installed it on an old X220i computer that has 256 SSD hard drive and 8 GB RAM, but it runs very well, no issues at all. Here are some stepwise notes on setting up a workable data science and analytical and productive system.

First, get the ISO installer from the following source, get the OpenSuse Tumbleweed version for your desktop or notebook, from here:

openSUSE Tumbleweed

The installer requires about 15 GB space and 2 GB RAM, but the more the merrier. Download it into a space in an existing computer (I used a Macbook or you can download on to another existing Linux box) using the following code:

sudo dd if=filename.iso of=/dev/sdb bs=128M

Here’s a short explanation of the above code:

• sudo is basically you as a superuser so the terminal will next ask you to enter your password
• dd is the command to burn the file into a disk. It will erase whatever existing data on the usb stick and install this iso
• if=filename.iso; if = input file, then name of the file
• of=/dev/sdb, here of = output file, and /dev/sdb is the name of the drive where the usb drive is located. You can find the correct drive address with either diskutil list on Mac or lsblk in linux box.
• For windows you may want to refer to the following link: https://windowsreport.com/burn-iso-files-windows-10/

Next, after you have burned the disk iso, pop it in the usb slot of the computer and reboot. Depending on the make of the notebook, you may need to identify the right key to enter the BIOS. After you have entered the bios, change the bootorder so that the computer will boot from the USB drive.

After this, let the computer boot and you will be brought to a screen that will ask you to install OpenSuse. Install it on your computer. I selected to erase all data on my existing hard disk and partition it. The partitioning system of OpenSuse is intuitive and it will do a guided partition and install a desktop environment. I selected a KDE-plasma environment that has a pleasing appearance and fast.

After this step, reboot the computer and it will bring you to a desktop where you can now install the rest of the system. This means you will need to install the following (some are optional but helpful):

• Install R for data analysis and graphics
• Install Julia (I use Julia for my work, but you may not use this)
• Quarto (a useful package to work with)
• CodeOSS or VSCode
• Jupyter Notebook, or specifically Jupyter Lab
• Connect everything

Installation of R is simple, open a console and type sudo zypper install R-base R-devel

What this code is doing:

1. sudo is about you being the superuse so be prepared to write your password as admin
2. zypper is OpenSuse’s package manager
3. install R-base R-devel are that you need these two packages to work and install is the installation command

Install Julia via juliaup, and it is really easy with OpenSuse as they have their own juliaup installer, so you do: sudo zypper install juliaup and then juliaup add release . Done.

Installation of Quarto is a little more involved, but here is the webpage that provides you with excellent guidance:

Quarto - Tarball Installation On Linux

Just follow or basically copy and paste the codes on the page and you are good to go.

Similarly, installation of vscode is as easy as it gets, basically you do sudo zypper install vscode and you are done.

For Jupyter Lab it is a little more involved:

1. First get micromamba with '$SHELL' < (curl -L micro.mamba.pm/install.sh)
2. What the above code does is that it downloads the install script and runs it in the bash shell. Make sure that you have curl installed, if not, install with sudo zypper install curl
3. Once micromamba is installed, create an environment with micromamba create -n myenv
4. The above code creates a new environment named myenv
5. Then activate this environment with micromamba activate myenv You will then see myenv activated in the shell
6. Then run micromamba install jupyterlab
7. It will install Jupyter environment with Jupyter lab

Now connect everything

If you do not connect the Jupyter Lab with the others, it won’t work. Likewise, for VSCode you need to connect and let the programme know where to find the respective software. Here are the steps for Jupyter notebooks:

• Install LanguageServer.jl package for Julia. Open Julia by typing julia in the command prompt, then press ] button to bring up the pkg module. In the pkg module, type add LanguageServer That’s it.
• Then after the LanguageServer module is installed, do add IJulia so that it can connect to Jupyter Notebook or Jupyter Lab
• For R, open an R console, start an R session, and then do install.packages("IRkernel") . In OpenSuse, it can end up with an error message that it did not install pdbzmq package. If you see this, start with sudo zypper install zeromq-devel This step is needed to install the zeromq-devel package. Then repeat the above step of installing IRkernel. Then after the IRkernel is installed, do IRkernel::installspec()
• Remember to do the above step WITHIN an R session. Otherwise this will not work.

Final Notes

These above steps will help you to set up a working data science environment in a computer running OpenSuse. You can use vscode to conduct your data analyses, alternatively, use Jupyter Lab for your data analysis and report writing and use Quarto for these activities. More on these in the next series of posts.