This is the fourth edition of the annual ESAC Data Analysis & Statistics (EDAS) Workshop. 

This four-day EDAS workshop will take place from Tuesday to Friday on November 28 to December 1 at the European Space Astronomy Centre near Madrid in Spain.

Two days will cover fundamental topics in statistics and data analysis, and two days will be for practical applications and more advanced topics. 

Each day of this year's EDAS workshop will be lead by a different tutor. 

The tutors for this year are Doctors Brendon Brewer, Ewan Cameron, Coryn Bailer-Jones, and Michelle Lochner (details below).




Registration closed on Friday 15 September 2017. Participants have been selected, and notifications have been sent. A contribution from external participants of 75 euros must be paid in cash on the first day of the workshop when picking up your name tag. Lunch tickets for 9 euros each (36 euros for 4 days) should also be purchased on the first day to simplify things. 


​Dr Brendon Brewer 

is a senior lecturer in statistics at the University of Auckland in New Zealand. His PhD was in astrophysics and he has worked on inference problems in gravitational lensing, reverberation mapping, asteroseismology, and cricket. His methodological interests are in computational Bayesian inference, with a focus on the Nested Sampling family of algorithms. (See also and Dr Brewer has compiled useful resources for students which are relevant for this workshop.



Dr Coryn Bailer-Jones   

is a senior staff member and group leader at the Max Planck Institute for Astronomy in Heidelberg. He is head of Coordination Unit 8 in the Gaia Data Processing and Analysis Consortium (DPAC). CU8 is concerned with source classification and the estimation of astrophysical parameters using the Gaia photometry, spectroscopy, and astrometry. His main scientific interests are stars, the Galaxy, and the impact of astronomical phenomena on the Earth. He also teaches physics and statistics at Heidelberg University. His book "Practical Bayesian Inference" was published in 2017 by Cambridge University Press. (See also



Dr Ewan Cameron   

is a senior computational statistician in the Malaria Atlas Project team at the University of Oxford. His work is focused on the development of novel methods for probabilistic map-making and the calibration of expensive simulation-based models for disease transmission, both of which are founded on fast and flexible techniques for Gaussian process regression. Dr Cameron has a background in the study of galaxy evolution and post-doctoral experience in the study of Bayesian methods applied to astronomical datasets. He also maintains an active interest in astrostatistics through his blog and his involvement with the Cosmostatistics Initiative. (See also and



Dr Michelle Lochner   

is a Resident Researcher at the African Institute for Mathematical Sciences (AIMS) and the Square Kilometre Array South Africa (SKA SA). She did her PhD at the University of Cape Town before moving to University College London for a two-year postdoc. She has now returned to South Africa to take up the position she currently holds. Her broad interest is in developing new techniques that enable precision cosmology, and she has worked on statistical and machine learning applications in the fields of supernova cosmology, radio interferometry and radio spectroscopy. As an active member of two of the most exciting upcoming astronomical experiments, LSST and the SKA, she is passionate about the role statistics and machine learning can play in extracting the most out of the incoming "mega data" we'll be seeing in the near future. (See also


AGENDA & timetable

schedule for each day

  • Start - 9:30 (ecxept the first day that starts at 9:00)
  • Coffee Break - 11:00 to 11:30
  • Lunch - 13:00 to 14:30
  • Coffee Break - 15:30 to 16:00
  • End - 17:30
Download detailed schedule here: EDAS.ics


Tuesday 28 November

The Bayesics. Session led by Dr Brendon Brewer, topics include: 

  • Probability theory
  • Probability distributions
  • Bayesian inference 
  • Metropolis algorithm
  • Nested sampling

The git repositories for Brendon's session can be found here: and

Slides are given here: 1. Probability2. Distributions3. Parameter Estimation4 Metropolis5 Nested Sampling6 Hierarchical models;

The associated exercises here: Questions 1Questions 2Questions 3Questions 4;

And the solutions to the exercises here: Solutions 1, Solutions 2, Solutions 3, Solutions 4.


Wednesday 29 November

Applications to GAIA. Session led by Dr Coryn Bailer-Jones, topics include:

  • Single objects - astrometry & photometry based parameter estimation
  • Multi-object - inter-source covariance, partial correlation, hierarchical modelling 
  • Priors from background information

The git repository for Coryn's session can be found here: (R) and (python).  

The file of the presentation slides is in the first repository and called astrostats_2017_ESAC.pdf (21 MB).

Coryn also wrote a short handout on the problem of infernce cluster distance that you can see here: cluster_inference.pdf


Thursday 30 NOVEMBER

Advanced Bayesian computation. Session led by Dr Ewan Cameron, topics include: 

  • Model design
  • Prior regularisation
  • Model checking
  • Approximate Bayesian inference

The slides for Ewan's session can be obtained here: esac_final.pdf (see also: pseudo_marginal.pdf, inla_example.pdf, sis_example.pdf).

You will also need the following files:, galaxy.R, pseudomarginal.R, tracking.R.


Friday 1 December

Machine Learning in astronomy. Session led by Michelle Lochner, topics include:

  • Introduction to machine learning
  • Machine Learning vs Statistical methods
  • Classification
  • Post processing 

The git repository for Michelle's session can be found here:


The workshop and hands-on sessions will be based on python and R.

Attendees are expected to come to the workshop with the following installed and running:

Video ​Lectures

  • Day 1 - Dr Brendon Brewer - video
  • Day 2 - Dr Coryn Bailer-Jones - video
  • Day 3 - Dr Ewan Cameron - video
  • Day 4 - Dr Michelle Lochner​ - video

All lectures are available on the ESAC Data Analysis and Statistics YouTube channel

We recommend the all-in-one scientific Python installer Anaconda. Download Anaconda from