.. _method-bayes_calibration-muq-multilevel_mcmc-greedy_resampling_factor: """""""""""""""""""""""" greedy_resampling_factor """""""""""""""""""""""" Greedy resampling factor **Topics** :ref:`bayesian_calibration ` **Specification** - *Alias:* None - *Arguments:* None - *JSON argument key:* None - *Default:* 0.5 **Description** MUQ's greedy algorithm frequently recomputes variance at each level. The ``greedy_resampling_factor`` sets the fraction of samples to reuse before computing. *Default Behavior* Defaults to 0.5. **Examples** .. tab:: Freeform .. code-block:: dakota method, bayes_calibration muq chain_samples = 1000 multilevel_mcmc initial_chain_samples = 50 target_variance = 0.5 subsampling_steps = 12 0 greedy_resampling_factor = 0.5 .. tab:: JSON .. code-block:: json { "method": [ { "bayes_calibration": { "sub_method": { "muq": { "chain_samples": 1000, "mcmc_algorithm": { "multilevel_mcmc": { "initial_chain_samples": 50, "target_variance": 0.5, "subsampling_steps": [12, 0], "greedy_resampling_factor": 0.5 } } } } } } ] }