sample_type

Selection of sampling strategy

Specification

  • Alias: None

  • Arguments: None

  • JSON argument key: None

Child Keywords:

Required/Optional

Description of Group

JSON Group Key

Dakota Keyword

Dakota Keyword Description

Optional

lhs

Uses Latin Hypercube Sampling (LHS) to sample variables

Optional

random

Uses purely random Monte Carlo sampling to sample variables

Optional

incremental_lhs

(Deprecated keyword) Augments an existing Latin Hypercube Sampling (LHS) study

Optional

incremental_random

(Deprecated keyword) Augments an existing random sampling study

Optional

low_discrepancy

Uses low-discrepancy points to sample variables

Description

The sample_type keyword allows the user to select between three types of sampling: Monte Carlo (pure random), Latin hypercube (stratified), and low-discrepancy (quasi-Monte Carlo) sampling.

The incremental keywords are deprecated; instead use samples together with refinement_samples.

Default Behavior

If the sample_type keyword is present, it must be accompanied by lhs, random or low_discrepancy. In most contexts, lhs is the default (exception: multilevel_sampling uses Monte Carlo by default).

Examples

method
  sampling
    sample_type lhs
    samples = 20
    seed = 83921
{
  "method": [
    {
      "sampling": {
        "sample_type": {
          "lhs": true
        },
        "samples": 20,
        "seed": 83921
      }
    }
  ]
}