Adaptive importance sampling

These tutorials will cover how to use the adaptive importance sampling method in COSMIC to efficiently sample from the parameter space and obtain results with fewer samples.

• Adaptive importance sampling for rare systems
  • When should I use this?
  • How it works: think Battleships
  • Setup
  • Running the sampler
  • Rules of thumb
  • Saving your results
• Distributions and custom priors
  • Built-in distributions
  • How distributions are built
  • Defining your own distribution
  • Custom transforms
  • Wrap-up
• Handling outputs from adaptive sampling
  • Reading your results from a file
  • Understanding your outputs
  • How to interpret adaptive sampling weights
  • Drawing a representative sample from your simulation
  • Wrap-up
• Saving and loading checkpoints
  • Why split a run in two?
  • Stage 1 — explore, adapt, and save a checkpoint
  • Stage 2 — load the checkpoint and refine
  • What is stored in a checkpoint
  • Reusing one checkpoint for several refinement jobs
  • Wrap-up