Fitting models with merlin

One command, a grammar, and the models it lets you write.

Taught by

Michael Crowther, PhD

CEO & Director of Statistical Methodology

Formerly Associate Professor of Biostatistics at the University of Leicester, and the author of the merlin framework.

Research survival analysis · joint models · multi-state models · statistical software

PhD in Medical Statistics, University of Leicester, 2014

Register your interest

What you will be able to do

• Read a merlin command and say what model it fits
• Fit a parametric survival model in merlin, and check it against a command you already trust
• Recognise when two commands disagree because of a parameterisation rather than a model
• Choose a baseline hazard, and put a spline on a covariate or on an effect
• Report standardised survival and a restricted mean survival time difference from a fit
• Say why the same model gives different numbers in Stata and in R, and which differences matter

Before you start

• Take first: Survival analysis
• Helps to have done: Flexible parametric survival models

Curriculum

27 steps: 18 lessons, 4 practicals and 5 quizzes.

Provisional — this course is still being written, and the outline will change.

Module 0 · Before you start

• 01ReadWhat merlin is, and what it is not8 min
• 02ReadInstall it, and prove it works10 min
• 03QuizFive questions before you start5 min

Module 1 · One model

• 04ReadThe model you already know12 min
• 05ReadThe same model in merlin15 min
• 06ReadReading the output, and why the log-likelihoods differ10 min
• 07ReadTwo parameterisations of the same Weibull10 min
• 08PracticalPractical 1 · Fit it twice35 min
• 09QuizWhat you just learned6 min

Module 2 · The right-hand side

• 10Readfamily() — choosing a baseline12 min
• 11ReadSplines on the baseline12 min
• 12ReadSplines on a covariate12 min
• 13ReadSplines on the effect — relaxing proportional hazards12 min
• 14BuildBuild the command15 min
• 15PracticalPractical 2 · Find the non-proportionality40 min
• 16QuizWhat you learned in module 26 min

Module 3 · What you actually report

• 17Readpredict, and the statistic menu12 min
• 18ReadAt what covariate values10 min
• 19ReadStandardised survival — the g-formula in one option14 min
• 20ReadDifferences and ratios10 min
• 21PracticalPractical 3 · Report an RMST difference40 min
• 22QuizWhat you learned in module 36 min

Module 4 · The second set of brackets

• 23ReadTwo models, one command14 min
• 24ReadWhat the brackets make possible12 min
• 25ReadWhere the two languages differ, and why12 min
• 26PracticalCapstone · One dataset, four answers45 min
• 27QuizThe whole course, in ten questions10 min