Analyses, experiments, and evaluations for the TrialMDP method.

See the TrialMDP manuscript on arXiv, or the journal article in Computational Statistics and Data Analysis. The Snakemake workflow in this repository reproduces all of the results described in that paper.

• Set up your R environment:

  • Install the TrialMDP package
  • Install the optparse package

• Set up your python 3.8 environment:

  • run pip install -r requirements.txt to install the requisite python packages

• Make sure you have sufficient compute resources.

  • The analyses entail hundreds of CPU-hours (can be parallelized by --cores option, see below)
  • The analyses require a few GB of memory and disk

• Call Snakemake.

  • On a workstation, execute snakemake --cores CORES (where CORES is replaced by the number of cores you want to devote)
  • On a cluster, execute snakemake --profile PROFILE (where PROFILE is replaced by some suitable Snakemake profile )