R for Health Data Research

Beginner, Intermediate and Advanced

Dr Ewan Carr

Department of Biostatistics & Health Informatics
King’s College London

Course materials 📚

All lectures and practicals can be found at the link below:

r-training.uk

• You should have received a password via email.

• Updated throughout the course and available for at least one month after the final session.

Submit your questions 🙋

Click the “Submit a question” link to send me questions during the week.

Pre-course videos and homework 📝

The sessions are spread over four weeks, allowing time for the material to sink in before moving on.

• Some sessions may have a short video to watch before the session.

  These will be posted on the HDR UK Futures platform.

• Some sessions contain extra ‘homework’ exercises. If you want more practice, have at go at these in your own time.

Learning a new programming language is hard.

Especially when you’re used to another way of working
(e.g., Stata, SPSS or Python).

I tried to learn R five times before it stuck.

Expectations 💡

• Expect confusion. It’s inevitable; you’re not doing anything wrong.
• Mistakes are normal, even after many years of using R.
• Allow time; things will take (much) longer at first.
• It can help to use R for a specific project, before switching completely.

After 16 years of using R, I still:

• Regularly make mistakes
• Have to look things up, check the documentation
• Spend forever tweaking code, and running endless tests, only to discover in the end that the entire problem was caused by a single comma out of place 😭

What this course is about ✅

• Building confidence in using R for health data science
• Learning to write, run, and understand R code
• Working with complex, real-world data — importing, tidying, transforming, and summarising
• Developing efficient, reproducible workflows in R
• Moving toward modern, professional programming practice

What this course is not about ❌

• Statistical modelling or inference
• Hypothesis tests, regression, or prediction modelling
• Detailed coverage of specific statistical packages

I’ll touch on these where relevant, but we won’t teach advanced statistics here.

This is a large, online class with a mix of lectures and practical sessions.

During the lectures 🗣️

• Post questions in the chat.

  (You can message me directly, or the whole class).

• Listen or type along — whatever works best for you.

During the practicals 💻

• Post questions in the chat or put a hand up.
• You’re welcome to unmute and share screen.
• I’ll use messages in the chat to check progress.
• I’ll go through the exercises at the end.

After the sessions 🕐

• Review material and lecture recordings.
• If you want, have a go at the homework.
• Send me your questions via the form.

Every question is helpful

Someone else is almost certainly
wondering the same thing.

Learning R is a journey

“If I take one more step, it’ll be the farthest away from home I’ve ever been.”

Coding with generative AI

Good uses ✅

• Quick help with debugging
• Explaining why code fails
• Generating toy examples or simulated data

Caution ⚠️

• When it affects real-world decisions (e.g., patient care, service delivery)
• When you don’t yet understand the problem
• Confident-sounding but wrong answers
• Outdated or odd R code (especially in fast-moving packages) Code you can’t explain is code you can’t maintain

Why I think we’ll still be coding in 5 years’ time

• As with statistics, the hard part is communication and judgement: framing the problem and choosing the right approach (not typing)
• Much of health data science is low-volume, high-importance work
• Analyses need to be traceable and reproducible
• Writing (and reading) code is part of the thinking
• Writing code is fun