matrix

Specifies the full matrix format when specifying a user-specified proposal covariance.

Specification

  • Alias: None

  • Arguments: None

  • JSON argument key: None

  • Default: true

Description

When specifying the MCMC proposal covariance in an input file, this keyword declares the use of a full matrix format, i.e., the user provides all values of the matrix, not just the diagonal. The matrix must be symmetric, positive-definite.

Examples

method,
        bayes_calibration queso
         samples = 1000 seed = 348
          dram
          proposal_covariance
            matrix values 1.0 0.1
                          0.1 2.0
{
  "method": [
    {
      "bayes_calibration": {
        "queso": {
          "samples": 1000,
          "seed": 348,
          "dram": {},
          "proposal_covariance": {
            "matrix": {
              "values": [1.0, 0.1, 0.1, 2.0]
            }
          }
        }
      }
    }
  ]
}