ToyMCMC

Purpose

ToyMCMC is an alias method that uses the baseline chain-sampling runtime.

Full Sampling Section Keys

• Sampling.Method (required): must be ToyMCMC.
• Sampling.Variables (required, array):
• name (required)
• description (required)
• distribution.type (required)
• distribution.parameters (required)
• runtime-safe parameter sets:
  • Flat: min, max
  • Log: min, max
  • Normal: mean, stddev
  • Log-Normal: mean, stddev
  • Logit: location, scale
• Sampling.LogLikelihood (required): array of {name, expression}
• Sampling.selection (optional, string)
• Sampling.Bounds (required):
• num_chains (integer)
• num_iters (integer)
• proposal_scale (number or array)

Full Skeleton

Sampling:
  Method: "ToyMCMC"
  Variables:
    - name: p1
      description: toy parameter
      distribution:
        type: Flat
        parameters:
          min: -5
          max: 5
  LogLikelihood:
    - name: L_total
      expression: "-0.5*(p1/1.5)^2"
  Bounds:
    num_chains: 4
    num_iters: 2000
    proposal_scale: 0.2

Example

Sampling:
  Method: "ToyMCMC"
  Variables:
    - name: x
      description: toy variable
      distribution:
        type: Flat
        parameters:
          min: -5
          max: 5
  LogLikelihood:
    - name: L_toy
      expression: "-0.5*(x/1.5)^2"
  Bounds:
    num_chains: 4
    num_iters: 2000
    proposal_scale: 0.2