Practical Inference for Researchers in the Physical Sciences

by Francesca Capel (ODSL), Johannes Buchner (PUC)



The ORIGINS Data Science Laboratory is organising the next set of block courses from September 6th - 15th 2021 on the theme of Practical Inference for Researchers in the Physical Sciences.

The idea with the block course format is to have two weeks of more intense lectures and tutorial sessions. With this course, we want to focus on bridging the gap between learning about Bayesian probability theory and the actual application of these methods to realistic research problems. In this way, the course will be a mix of standard lectures as well as interactive coding exercises.

The courses and tutorials are organised by Johannes Buchner and Francesca Capel. All sessions will be held online via Zoom. Please register for the course using the link below and feel free to get in touch if you have any questions.

Course content

This session consists of two one-week courses.

Block I: Monte Carlo inference methods

September 6th - 8th 2021

In this block, we will refresh some of the basic concepts of Bayesian inference and introduce the algorithms and tools that can be used to implement analyses. The main lecturer is Johannes Buchner. The course content can be found in this GitHub repository.

Topics covered:

  • Refresher on Bayesian inference for parameter estimation and model comparison
  • Parameter uncertainties, degeneracies and knowledge updates
  • Modern Monte Carlo algorithms for Bayesian inference in practice:
    • Importance Sampling
    • Markov Chain Monte Carlo
    • Nested Sampling
  • Modern probabilistic computation packages

Block II: Bayesian workflow

September 13th - 15th 2021

In this block, we will demonstrate how to apply the above methods to realistic problems and deal with common issues that may arise. The main lecturer is Francesca Capel. The course content can be found in this GitHub repository.

Topics covered:

  • Going from a science question to a statistical model
  • Defining sensible priors for your problem
  • Diagnosing problems in models and computation
  • Verification of a statistical model through simulations
  • Experiment design
  • Model comparison

Expected prerequisites

We expect students following the course to have a basic understanding of Bayesian probability theory and a good grasp of python and Jupyter notebooks, which will be used for the tutorial sessions.

Start here: Enter the course, install software


The course runs from Monday to Wednesday each week with the following schedule:

Morning session 10:00 - 12:00 

Afternoon session 14:00 - 16:00

These sessions are a mix of lectures and interactive coding tutorials. We will include some short breaks within the 2 hours.


To register, click on the button below. If you want to register for ECTS credits through the TUM, you must also register in TUMonline (simply search for each block course by its name). 

Report and forms of credit

The course content includes exercises, homework exercises and a homework project which can be completed for points. We also encourage your active participation in the course, and further points for speaking up and interacting in the zoom sessions.

Total points available for each block course:

  • Finding mistakes: 5 
  • Active in class: 50
  • Exercises: 150
  • Homework exercises: 300
  • Homework project: 200 

Outside of class, points will be counted based on a report to be submitted for each block course in .pdf format. The report should clearly state the problems you are attempting, along with code snippets and plots to show your working and solutions. The deadline to hand in your report is September 30, 2021. 

It is possible to get a certification or ECTS points for participation in the block courses:

To get a Certificate of Participation, you need at least 100 points in each block course (i.e. at least 100 points in the first block and 100 points in the second block course).  The certification will state that you have successfully completed the block course in the respective topic.  

To get the 3 (LMU) ECTS points, you need at least 300 points in each block course (i.e. 300 points in the first and 300 in the second block). 

To get the 5 (TUM) ECTS points, you need at least 500 points in each block course (i.e. 500 points in the first and 500 in the second block).

Participants who do not want a certificate or ECTS credits are not required to turn in a report, but are allowed to do so.

Receiving ECTS via the TUM or LMU

TUM: Students may obtain 5 ECTS by registering on TUMOnline (see Registration above).

LMU: Students in Physics may obtain 3 ECTS as an elective subject from
this course. Please follow the directions at
for a recognition of the credits.

Zoom information

The Zoom meeting link for Block I will be shared after this questionnaire prior to the start of the course.

The Zoom meeting link for Block II is:

Meeting ID: 693 6960 5092

Passcode: 357226


The course lectures will be recorded. If you have any objections to this, please contact us directly. We will make the recordings of the lectures available online (see below).

Lecture recordings

Block I: At this link

Block II: At this link




Organized by

Francesca Capel
Johannes Buchner

Course evaluation