Available Benchmarks

Browse Benchmarks

Below you’ll find all available benchmark datasets. Each benchmark includes detailed documentation, data files, and associated modeling tasks.

Using the Benchmarks

All benchmarks are freely available under their specified licenses. You can:

  • Download datasets directly from this repository
  • Use helper packages (coming soon) to programmatically access benchmarks:
    • Python: pmx-benchmarks-py
    • R: pmxbenchmarks
    • Julia: PMXBenchmarks.jl
  • Cite benchmarks in your publications using their DOIs

Benchmarks

Example PK Model Selection

A synthetic two-compartment pharmacokinetic dataset for evaluating model selection methodologies.

  • Authors: Jane Researcher, John Modeler
  • Data Type: Synthetic
  • Subjects: 200 (140 train, 60 test)
  • Tasks: Structural model selection, covariate selection, prediction accuracy

View Full Documentation | Download Data

More benchmarks will appear here as they are submitted and accepted. Want to contribute? See our submission guide.