.. _method-multifidelity_polynomial_chaos-statistics_mode:

"""""""""""""""
statistics_mode
"""""""""""""""


type of statistical metric roll-up for multifidelity UQ methods


.. toctree::
   :hidden:
   :maxdepth: 1

   method-multifidelity_polynomial_chaos-statistics_mode-active
   method-multifidelity_polynomial_chaos-statistics_mode-combined


**Specification**

- *Alias:* None

- *Arguments:* None

- *Default:* DEFAULT_EXPANSION_STATS


**Child Keywords:**

+-------------------------+--------------------+--------------------+-----------------------------------------------+
| Required/Optional       | Description of     | Dakota Keyword     | Dakota Keyword Description                    |
|                         | Group              |                    |                                               |
+=========================+====================+====================+===============================================+
| Required (Choose One)   | Statistics mode    | `active`__         | use statistics for the active expansion       |
|                         | for refinement     |                    | within multifidelity refinement               |
|                         | metric             +--------------------+-----------------------------------------------+
|                         |                    | `combined`__       | use statistics for the combined expansion     |
|                         |                    |                    | within multifidelity refinement               |
+-------------------------+--------------------+--------------------+-----------------------------------------------+

.. __: method-multifidelity_polynomial_chaos-statistics_mode-active.html
__ method-multifidelity_polynomial_chaos-statistics_mode-combined.html



**Description**


When performing adaptive refinement in multifidelity UQ
methods, one can adapt the expansion for a model level/fidelity based
on its local statistics ( ``active``) or based on the statistics for the
aggregate expansion ( ``combined``).  The latter case may incur
additional computational overhead, but ensures that the refinement
occurs in the aggregate multifidelity context.

*Default Behavior*

Default for multifidelity methods is ``combined`` statistics.

*Expected Output*

*Usage Tips*

*Additional Discussion*