# Software and flavors

Your server's software comes in packages we call **flavors**: a programming language with its libraries, a toolkit, or an application. Your lecturer or project lead chooses the flavors of each environment, so a Python course gets Python and nothing it does not need.

## What you have

The flavors of your environment are ready as soon as your server starts:

- their **kernels** appear in the JupyterLab Launcher, for example *Python 3.11 (Scientific)* or *Julia*;
- their **programs** work in the terminal;
- graphical applications such as MATLAB or QGIS appear in the **desktop** menu.

The environment card on the start form lists the flavors of each environment.

## Adding software during a session

**Another flavor.** Every flavor is installed on the machine, even those your environment does not include. Load one in a terminal:

```bash
module avail              # every flavor available on this machine
module load matlab        # make MATLAB available in this terminal
module list               # what is loaded right now
```

**A Python package.** Install it into your home folder:

```bash
pip install --user <package>
```

It is kept between sessions and works in new notebooks without restarting the server.

**Something a whole course needs.** Ask your lecturer to add the flavor to the environment, or, if it is not in the catalog below, [ask us](https://datalab.tuwien.ac.at/science/jaas/#help-and-news).

## Flavor catalog

| Flavor | Contents | Category |
|---|---|---|
| **Scientific Python + R** | Python 3.11 and R with the scientific stack, from PyTorch and JAX to tidyverse and Bioconductor | Languages |
| **Julia** | Julia 1.11 with the dataLAB package depot: Plots, Turing, PyCall, Pluto | Languages |
| **C / C++** | C and C++17 Jupyter kernels (xeus-cling) | Languages |
| **MATLAB** | MATLAB R2024b for notebooks and the desktop | Applications |
| **GAMS** | General Algebraic Modeling System 24.7 (licensed) | Applications |
| **Geodesy / LiDAR** | OPALS, CloudCompare and QGIS for airborne laser scanning, point clouds and GIS | Applications |
| **CUDA toolkit** | nvcc, ncu, the CUDA libraries and Nsight Systems for compiling and profiling GPU code | Tools |
| **TeX Live** | XeLaTeX toolchain for exporting notebooks to PDF | Tools |

The **Grader** extension for [assignments](https://datalab.tuwien.ac.at/science/jaas/assignments/) is a flavor too. It is added to every course automatically and cannot be selected.

## Versions

Flavors are versioned, for example `sci/2026.07`. A new version is published next to the old one, so a running course keeps its version until the lecturer switches. You get the new version the next time your server starts.

## Next steps

- [Working in a group](https://datalab.tuwien.ac.at/science/jaas/group-work/)
- Lecturers: [choosing flavors for an environment](https://datalab.tuwien.ac.at/science/jaas/instructors/environments/#software)