Embedded Files
Munich, Germany
5-day workshop on reproducible computational research
5-day workshop on reproducible computational research
Join us for our annual, intensive 5-day workshop on computational reproducibility and scientific computing, organized by the MCML. Each day runs from 9:00 AM to approximately 5:00 PM and combines lecture-style sessions with hands-on practical work. Compared with the previous two editions, the program is being substantially updated to account for rapid developments in agentic AI. The preliminary program, which may still be subject to change, will provide a condensed introduction to core topics such as reproducible manuscripts, high-performance computing (HPC), workflow management, containers, testing, packaging, and CI/CD during the first three days. The final two days will focus on practical tutorials using coding agents for scientific computing and reproducible research workflows.
Participants' feedback
Participants' feedback
"Really happy overall with the workshop. Many advanced concept that are also useful in the daily work, and that are hard to get the time to get a proper idea."
"Thanks for this great workshop! I will definitely recommend it to other colleagues :) I hope you will provide it again in the future."
"Very relevant topics, great hands-on sessions, helpful slides. Thank you very much for making the workshop possible."
I really loved the workshop! I actually wish it was longer because it was very dense for 5 days. I learned so much and I’m happy that there is a workshop where reproducibility is discussed. I can see that the organizers have put in a lot of thought and hardwork behind the materials, so thank you! :-)
Details of the workshop
Who: Anybody interested, the primary target group is junior MCML PhD students
(Max. 60 participants, please see prerequisites and registration)
Where: Munich, in-person, Location: Mottlestr 1, 80804 Munich, Room: TBD
What: Topics include reproducible manuscripts, HPC, Packaging, CI/CD, agentic AI & coding agents, both for R and Python users
When: Mo-Fr, 05. - 09.10.
If it is organizationally feasible, we will try to enable participation on a per-day basis.
Mo, 05.10.
Mo, 05.10.
Day 1: Foundations of Computational Reproducibility
Day 1: Foundations of Computational Reproducibility
Welcome and introduction to WS
Open science, reproducibility, and the replication crisis
Computational reproducibility in the age of agentic AI
Project-orientation, reproducible analysis & manuscripts
Tu, 06.10.
Tu, 06.10.
Day 2: Reproducible Computing at Scale
Day 2: Reproducible Computing at Scale
High-performance computing: Slurm and LRZ
Workflow management
Containerization
Reproducible and safe execution on remote infrastructure
We, 07.10.
We, 07.10.
Day 3: Research Software Engineering
Day 3: Research Software Engineering
Modular and reusable research software
Documentation and dependency management
Testing
Packaging, CI/CD, licensing, and MLOps
Th, 08.10.
Th, 08.10.
Day 4: Agentic Scientific Computing I
Day 4: Agentic Scientific Computing I
Primarily hands-on sessions
Introduction to coding agents and agentic workflows
Specifying tasks, constraints, and acceptance criteria
Safe and controlled use of coding agents
Auditing and improving research repositories with coding agents
Fr, 09.10.
Fr, 09.10.
Day 5: Agentic Scientific Computing II
Day 5: Agentic Scientific Computing II
Primarily hands-on sessions
End-to-end agent-assisted scientific workflows
Reproducing, testing, and documenting computational analyses
Preparing workflows for containers and HPC
Verification and critical assessment of agent-generated results
Prerequisites
Prerequisites
This year's workshop will be more language-agnostic than the previous versions. Nevertheless, we expect participants to be able to program in R or Python and use Git to make the most out of the workshop.
In general, we consider MCML PhD students to fulfill the requirements.
If you are a student of the Department of Statistics at the LMU and an R user, you should have finished the Introduction to Statistical Software ("StatSoft") and Programming with Statistical Software ("ProgR") courses. We refer other R users to this list.
Python users should have a level of programming skills that is equivalent to the topics covered in this course: Data Science with Python
You must bring your own laptop! Depending on the practicals you wish to participate in, we expect you to have a suitable version of Python and/or R, along with any other dependencies, installed. We will give a more detailed list of requirements before the workshop.
Participation / Registration
Participation / Registration
Participation is open to all members of LMU and TUM as well as students and researchers from other institutions.
Seats will be given on a first-register, first-serve basis. Half of the seats are reserved for members of the MCML.
Please register for the workshop by 18.09.2026 by filling in this form.
Contact
Contact
If you have any questions, please reach out to moritz.herrmann@lmu.de or ma.binder@lmu.de.
Moritz Herrmann
Moritz Herrmann
MCML Reproducibility and Open Science Transfer Coordinator
Martin Binder
Martin Binder
MCML Open Source and Open Data Transfer Coordinator
Page updated
Google Sites
Report abuse