Red Door Analytics

Most of the difficulty in survival analysis is not fitting the model. It is knowing which one the question calls for, and what you have assumed by choosing it.

This ends somewhere specific every time: a named model, the package that fits it, the syntax in both languages, and a tutorial on this site that works it through on real data. It covers the models the merlin family covers. If your design falls outside them it says so rather than offering the nearest thing that runs.

Notes on the methods behind this

Occasional notes on the methods behind these tools, new releases of merlin, avalon and pendragon, and courses. Monthly, roughly.

No tracking pixels; unsubscribe at any time. See our privacy policy.

Once you know the model

The model builder takes it from here: pick the baseline, the clustering and the extensions, and it writes the command. The glossary defines every term either page uses, and the test suite sets out what each family has been checked against.