This repository contains the code and TeX sources for the manuscript
`A review of Bayesian perspectives on sample size calculation for confirmatory trials'.
The R code to reproduce the figures is contained in the Jupyter notebook
notebooks/figures-for-manuscript.ipynb.
The TeX source files are contained in the latex subfolder.
All software dependencies are documented in the .binder subfolder.
The repository is compatible with the python tool repo2docker.
This means that the repository can be turned into a docker container that contains all
required software using repo2docker.
Using this container to execute the code then ensures a maximum level of
reproducibility.
We also use the (free) service mybinder.org to provide and interactive version of the repository online that can be explored without installing any software on your local system. The most recent development version can be accessed by clocking on the badge below.
Older releases of the repository are available via the badges listed below.
A Snakefile is provided to run all steps required to produce the final manuscript PDF file. To execute the workflow, simply open a shell (locally or via the mybinder.org link), and run the command
snakemake --cores 1 manuscript
in the project root directory (requires snakemake, ore-installed in the binder container).
Older releases:
Simply click on the badge to launch a simple shiny app and explore the sample size calculation rules interactively.