Programming Café

Do you have a computational workflow that consists of multiple interdependent steps (e.g. in data processing, analysis or modelling)? Do you want to automate your workflow and make your workflow reproducible and scalable? Then Snakemake might be interesting for you! 

Snakemake helps to automate your workflow in a reproducible way. It can connect to HPC and storage systems (such as YoDa) and allows you to run parts of your workflow in parallel with little effort. 

In this session of the Programming Café, we explored Snakemake, how it works and how to implement Snakemake in your project, whether you're using Python, R, Julia or other languages. Snakemake is based on Python, but no prior Python knowledge is necessary. Links here to install conda and Snakemake.

Marimo is a new, modern notebook environment designed to make your work reproducible by design. Unlike traditional notebooks, its reactive execution model keeps everything in sync, so you never have to wonder whether your results match your code. Marimo plays nicely with Git, making version control painless. It also lets you add sliders and other interactive UI elements to visually explore your data.

In this session of the Programming Café, we got hands-on with Marimo: exploring its features and converting existing Jupyter notebooks into Marimo notebooks. Participants left with cleaner, more reproducible, and more interactive workflows ☕🐍 Link to Marimo for researchers.

Some love it, some hate it, but every programmer regularly needs to do it: finding and solving the mistakes in your own program code. 

Or, if you are (un)lucky, in someone else's program code. Julian Gonggrijp (developer at the Research Software Lab, Digital Humanities) has been doing this professionally for twelve years and reached level 10. After briefly clarifying what that means, he shares his tools in a live debugging session.

Participants were invited to send a Python script of 300 lines or less with a tough bug in advance for the demo.

The large package ecosystem in R makes the language an appealing choice for may scientists. While packaging code is best practice in open science, it often requires investment of (scarce) time and a good understanding of the R ecosystem.

In this programming café, Niklas Hohmann shared why writing and publishing R packages is useful for researchers of all career stages and went through some quality-of-life setups that simplify package development. A main focus was on package development and publication according to the FAIR principles (Findable, Accessible, Interoperable, Reusable), and how these fit within the broader open science movement.

You can still come by during our walk-in hours for personalized support from our research engineers. That's why we're here! To request a new programming cafe on one of these themes, or have a consultation about your research project, email info.rdm@uu.nl, or fill in this form to request research engineering support directly.