multiresolution_multirun_table

multiresolution_multirun_table … end_multiresolution_multirun_table

multiresolution_multirun_table
   model_list
      <platform> <component> [<component-name>]
      <platform> <component> [<component-name>]
      ...
   end_model_list

   default_fidelity
      <real-value>
      <real-value>
      ...
   end_default_fidelity

   fidelity_table
      <real-value> <real-value> ...
      <real-value> <real-value> ...
      ...
   end_fidelity_table

   loop_after_table_end

end_multiresolution_multirun_table

Overview

multiresolution_multirun_table is a structure that allows a user to specify a number of fidelity values for a list of multiresolution models (e.g., multiresolution_mover). The table works in conjunction with the monte carlo iteration implementation, using the same internal RunNumber variable in order to select the row of fidelity_table.

The Python3 module pyrunplotter (located in the tools directory) can be used to visualize and compare the outputs from multiple runs.

Commands

model_list

Contains a list of models to be used. The models are structured <platform-name> <component-type> for unnamed components and <platform-name> <component-type> <component-name> for named components.

default_fidelity

Contains a list of fidelity values to use as defaults on the list of models. If the number of runs exceeds the number of rows on fidelity_table and looping is disabled, the simulation will fall back on these default values.

Note

If loop_after_table_end is enabled, this command can be omitted.

fidelity_table

Contains a MxN table of <real-value> entries, where M is the number of models in model_list and N is the desired number of runs. Each row in this table corresponds to a run, and each column corresponds to the model which will be assigned the column’s entries.

loop_after_table_end

Tells the simulation whether to loop through the entries in fidelity_table, or to fall back on the entries in default_fidelity when the number of runs exceeds the number of rows in fidelity_table. If enabled, this allows users to do multiple runs of each row, which may be desired if the users want to also use randomized variables for other aspects of the simulation.