Skip to contents

Learn R for clinical work, by doing it

R for Clinical Statistics is a gentle, hands-on course for clinicians and clinical researchers. You write real code against real-feeling clinical data from the first lesson, and every step gives you instant, kind feedback. No prior programming needed.

A course exercise: type real R code, run it in a safe sandbox, and get instant feedback.
A course exercise: type real R code, run it in a safe sandbox, and get instant feedback.

Why this course

The idea is simple: do something real in R within minutes, then build from there. Three promises hold in every lesson:

  • You cannot break anything. Each exercise is a safe sandbox. Run it, get it wrong, run it again.
  • Red text is not failure. An error is R asking for a small change, nothing more.
  • You are never stuck. A hint and a full solution sit one click away.

What you will be able to do

By the end of the course you can take a clinical dataset from a raw file to a result you would put in a paper:

  • pick the rows and columns you want, count groups, and sort (filter, select, count, arrange)
  • build new columns and summarise by treatment group (mutate, group_by, summarise)
  • reshape between long and wide, and join tables together (pivot_wider, left_join)
  • load and clean real data: inconsistent categories, numbers stored as text, missing values
  • draw the clinical figures a paper needs, and build a “Table 1” in one line
  • describe a variable properly, run a t-test or chi-squared test, and tidy the result
  • fit a linear or logistic regression, with odds ratios you can report
  • run a Kaplan-Meier and Cox survival analysis, censoring included
  • write the whole analysis up as a Quarto report that rebuilds itself from the data on demand

The lessons

The course is a set of interactive tutorials you run on your own computer.

Part 1 · Foundations

  • 0 · Why R for clinical data — what a script buys you, and your first line of R.
  • 1 · Tidy data as design — the one idea about table shape that makes everything else easy.
  • 2 · The building blocks of R — objects, value types, and what a data frame is.
  • 3 · Wrangling I — filter, count, arrange, select, and the pipe.
  • 4 · Wrangling II — mutate, and grouped summaries.
  • 5 · Tidy data in practice — reshaping and joining repeated measurements.

Part 2 · Real data

  • 6 · Reading & cleaning data — loading a file, and fixing the three messes you meet most.
  • 7 · Visualisation I — the ggplot2 grammar: scatter, box and bar plots.
  • 8 · Visualisation II — trends over time, faceting, and figures ready for a paper.
  • 9 · Summary tables — the “Table 1” every clinical paper opens with, via finalfit.

Part 3 · Analysis

  • 10 · Describing data — histograms first, then choosing mean/SD or median/IQR.
  • 11 · Comparing groups — the t-test, the chi-squared test, and tidy results with broom.
  • 12 · Regression — linear and logistic models, baseline adjustment, odds ratios.
  • 13 · Survival analysis — censoring, Kaplan-Meier curves, log-rank, Cox hazard ratios.

Part 4 · Reporting

  • 14 · Reproducible reports — the capstone: the whole trial analysis as a Quarto report, rendered to a real document on your machine.

Each interactive lesson ends with open practice exercises, graded as you go.

Get started in three steps

  1. Install R and RStudio, both free.
  2. Install the course package.
  3. Run learn() to see the lessons, then learn("m0") to open the first.

The Get started guide walks through each step with the exact commands. In a hurry? Try R in your browser first, with nothing to install.

Going deeper

When you want to read more, these four are worth your time: