.. _method-multifidelity_function_train-p_refinement-uniform-increment_max_rank_order:

""""""""""""""""""""""""
increment_max_rank_order
""""""""""""""""""""""""


candidate generation by advancement of maximum rank and maximum basis order


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



**Specification**

- *Alias:* None

- *Arguments:* None


**Description**


Within adaptive refinement processes, this specification selects the
advancement of both ``max_rank`` and ``max_order`` to create new
refinement candidates.  This option requires the use of ``adapt_rank``
and ``adapt_order``, for which this advancement creates expanded ranges
for the cross validation (CV) process.  Once the CV process saturates
for one or more of these two dimensions (the best rank is less than the
max rank and/or the best order is less than the max order), then no
further advancement can occur in that dimension for a particular
model level/fidelity.


