Small-group training and one-to-one coaching for data visualisation and Bayesian analysis.


Future courses

Advanced Statistical Analysis with Understanding ModernGov, 19 June 2019


19 June 2019
Advanced Statistical Analysis: improve performance through advanced statistical methods


A highly developed understanding of using statistical models can help your organisation monitor and improve its performance.
This Advanced Statistical Analysis course has been specially designed to help those with a background in statistics to understand and use more advanced statistical models, to make better sense of their data.
By attending this Advanced Statistical Analysis course, gain a better understanding of the concepts behind advanced statistics; learn how to apply the Bayesian model and utilise more complex statistical methods.
You will practice using R programming on public sector datasets, which will help you gain a firmer understanding of more innovative statistical methods.

Book here


Statistical analysis for clinical audit: one-day workshop in London, 21 June 2019



21 June 2019
Statistical analysis for clinical audit: a one-day workshop


This course is a one-day overview of statistical analysis applied to clinical audit. It is aimed at local audit teams as well as national projects. It will not turn you into an expert statistician but it will demystify the terminology and the techniques that are used, and give you an understanding of what can be done, when it might be useful, and what requires expert input or specialist software.

Advice and training in analysis for clinical audit tends to cover only basics, but increasingly, audits are expected to do more. In order to draw meaningful conclusions about performance, it is essential to deal correctly with problems like missing data, casemix differences, or the effects of external factors like season.

Even if your team would outsource this work, it is important to understand enough to make sure the deliverables are realistic and the subcontractors have the track record to achieve them.

This course is practical and highly interactive. It focuses on concepts rather than mathematics or software. No prior training in statistics is needed! It will be led by Robert Grant, a medical statistician of 20+ years' experience, who analysed many audits at the Royal College of Physicians before joining the Department of Health committee for clinical audit (NCAAG, later NAGCAE), providing methodological advice. He has taught statistics at St George's Hospital, Harvard Dubai Foundation, Kingston University, for many clients and online. His particular interest is in making the subject accessible and useful to less technically-trained audiences.

Learning outcomes - after taking this course, participants will understand:

  • how statistical models and related techniques can provide casemix adjustment, seasonal adjustment
  • how visualising audit data can help to detect errors as well as communicate results
  • how missing data can be addressed and what expertise is needed to do it
  • why care needs to be taken to pre-specify the analysis plan
  • how to make fair comparisons between units or time periods
  • what role machine learning techniques could play in their work
  • how to make a value-for-money choice of software resources to carry out audit analysis

Book here

Introduction to Bayesian Analysis Using Stan: Royal Statistical Society, 9-10 July 2019


9-10 July 2019
Introduction to Bayesian Analysis Using Stan


This two-day course is ideal for beginners or intermediate users of Bayesian modelling, who want to learn how to use Stan software within R (the material we cover can easily be applied to other Stan interfaces, such as Python or Julia). We will learn about constructing a Bayesian model in a flexible and transparent way, and the benefits of using a probabilistic programming language for this. The language in question, Stan, provides the fastest and most stable algorithms available today for fitting your model to your data. Participants will get lots of hands-on practice with real-life data, and lots of discussion time. We will also look at ways of validating, critiquing and improving your models.

Book here


From Statistics To Machine Learning, 28 October 2019


28 October 2019
From Statistics To Machine Learning

(early 2020 date to be confirmed: email robert@bayescamp.com to reserve a seat)


This one-day workshop on 28 October is aimed at anyone who studied some statistics in the past, and wants to understand the principles of machine learning. There are a number of techniques and ways of thinking that can be useful in any form of data analysis.

We will combine discussions about theory and working practices with thought-provoking small-group exercises. You will learn about:

  • Terminology and jargon
  • Supervised and unsupervised learning
  • Ensembles, bagging and boosting
  • Neural networks, image data and adversarial thinking
  • AI and ethical concerns
  • Reinforcement and imitation learning
  • Big data's challenges, opportunities and hype
  • Speed and memory efficiency
  • Concepts of model building such as cross-validation and feature engineering
  • Options for software, outsourcing and software-as-a-service
  • Data science workplaces combining statistical expertise with machine learning: what makes them happy and healthy

You can bring a laptop to try out some of the examples in R, but this is not essential. Refreshments and lunch will be provided.

Robert Grant is a medical statistician by training, more recently involved in machine learning techniques, who runs his own training and coaching company, BayesCamp. His specialities are Bayesian modelling (he is one of the contributing developers of Stan) and data visualisation (his book "Data Visualization: charts, maps and interactive graphics" is on CRC Press). He is currently test-driving around 20 different commercial machine learning software packages with the aim of publishing reviews and comparisons. He has many years' experience of teaching introductory courses and is committed to making advanced data analysis accessible to everyone who's interested.



Book here


Bayesian Meta-analysis, 4 November 2019


4 November 2019
Bayesian meta-analysis


This one-day workshop will introduce you to Bayesian models that solve some very widespread problems in meta-analysis:

  • unreported stats
  • some studies report means, others medians
  • some studies have dichotomised the outcomes
  • some studies are suspected of reporting bias or p-hacking
  • some studies are randomised, others are observational
  • studies have different intervention durations and follow-up
  • times
  • a mixture of different outcome measures and scales

We will use BUGS, JAGS and Stan within R. You don't need to be an expert in any of these, but you should at least have some familiarity with basic modeling in one of them. You will leave with tested code that you can use on your own data, and we will discuss the challenges that attendees have encountered with their own meta-analyses. The techniques covered here are applicable to many fields, including healthcare, econometrics and policy evaluation.

Facilitator Robert Grant has worked on meta-analysis and Bayesian models for many years, having been part of the NICE guideline development technical staff from 2000-2006. He is responsible for innovating several of the techniques we will cover today.



Book here


Get Up and Running in Bayesian Data Analysis with Rasmus Bååth, 25 Nov 2019


25 November 2019
Get Up and Running in Bayesian Data Analysis with Rasmus Bååth

25 November 2019, Central London TBC, £270
(£125 for students / unwaged -- email robert@bayescamp.com for a discount ticket)

In this one-day course for beginners in Bayes, you will learn how to think about problems in a Bayesian framework and the many advantages this brings, before getting some real experience of using R or Python packages, JAGS and Stan to carry out the computation. There is no need for prior experience of Bayesian analysis. You just need to have some familiarity with R or Python.

The course will be taught by Rasmus Bååth, a popular blogger and YouTuber on Bayesian methods, based at Lund University and King (makers of Candy Crush). Rasmus is particularly interested in making Bayes acessible to beginners and has experience through academic teaching, writing and creating online courses.
Book here

Stata courses with Timberlake: online and in London


1 November 2019
Introduction to Survival Analysis


Survival, or time-to-event, analysis has been widely used in biomedical research for decades but is also increasingly relevant to analysts working in fields like marketing, social sciences and engineering. Data follow people, organisations or other subjects over time and capture the times at which an event of interest happens. A cancer researcher may be watching for recurrence of a tumour, while in marketing, the event of interest could be a customer calling in to cancel their contract.
Stata has an established and trusted capability to manipulate such data and carry out a wide range of analyses. This half-day online course, delivered by experienced medical statistician Robert Grant, will introduce newcomers to the tools that Stata offers.
Topics covered:

  • Why survival / time-to-event data requires specific analyses
  • Defining and manipulating a survival / time-to-event dataset in Stata
  • Obtaining rates, survival and hazard curves from Stata
  • Testing for differences between groups in Stata
  • Adjusting for confounding variables in tests of hazard difference by stratification, using Stata
  • The role of regression in survival analysis, the difference between parametric and semi-parametric (Cox) models, and the basics of how to fit them in Stata

Contact Timberlake to book



23 September 2019
Stata Autumn School, day 1:
Introduction to Bayesian analysis in Stata


Stata has rapidly built a selection of Bayesian data analysis tools since version 14. The release of version 16 (anticipated summer 2019) adds more to this. There are also existing community-contributed interfaces to WinBUGS, JAGS and Stan. This day at Stata Autumn School will introduce these options and give participants a grounding in what Bayesian analysis is and how it can be useful. Further details of material to be covered will be added following the release of version 16.
Book here



24 September 2019
Stata Autumn School, day 2:
Multivariate analysis and unsupervised learning in Stata


Finding patterns in datasets of many variables is a key skill in contemporary data science, whether you call it multivariate analysis or unsupervised learning. This day at Stata Autumn School will look at the many options available in Stata and give participants a grounding in choosing an appropriate method and critiquing results.
Topics covered:

  • Cluster analysis
  • Principal components analysis
  • Factor analysis and the link to structural equation models
  • Correspondence analysis
  • Procrustes analysis of shapes
  • Options for data visualisation

Book here



18-19 November 2019
Bayesian data analysis in Stata 16


TBC Central London

Stata has rapidly built a selection of Bayesian data analysis tools since version 14. The release of version 16 (anticipated summer 2019) adds more to this. There are also existing community-contributed interfaces to WinBUGS, JAGS and Stan. This two-day workshop is for beginners in Bayesian methods, who want to understand how to fit models to their data and draw flexible inferences, all using Stata. Presenter Robert Grant is an experienced Stata user, trainer and medical statistician, who previously contributed the StataStan interface to Stan. Further details of material to be covered will be added following the release of version 16.
Contact Timberlake to book


Data Visualization with Physalia Courses, 9-13 Dec 2019


9-13 December 2019
Data Visualization

The ability to visualise data and analyses effectively is a highly prized skill in the data analysis workplace. Increasingly, employers are placing a high value on the ability to communicate as well as calculate. In this week-long course, you will learn how to design visualisations such as charts, maps and interactive graphics, which bring insight and understanding to your audience. You will learn about perception, the design process and the tools that you can use to make visualising easy. The course is highly interactive with group exercises throughout.
There will be two versions: for those who wish to practice some coding, for example with R, and for those who wish to work on paper throughout. In the last two days, participants in their groups will carry out a larger "project", where we present real-life data with a briefing from a researcher. The challenge will be to produce some visualisations and pitch them to the researcher on the last day, with feedback on your project work.
After taking this course, you will:

  • understand the options in converting data to an image
  • have practised sketching and user-testing the core of the data visualisation process
  • understand perception and what we can learn from it
  • have encountered a range of statistical challenges and their visualisation needs
  • have had the chance to try out several visualisation tools

Book here


Winter / spring courses to be confirmed soon:


You can also contact Robert at robert@bayescamp.com to discuss bespoke training for your team.



BayesCamp Ltd is incorporated in England and Wales, number 10666858. The term 'BayesCamp' and the Gaussian tent logo are registered trademarks.