CPLEX Optimization node (SPSS Modeler)

CPLEX Optimization node

With the CPLEX Optimization node, you can use complex mathematical (CPLEX) based optimization via an Optimization Programming Language (OPL) model file.

For more information about CPLEX optimization and OPL, see the IBM ILOG CPLEX Optimization Studio documentation.

When outputting the data generated by the CPLEX Optimization node, you can output the original data from the data sources together as single indexes, or as multiple dimensional indexes of the result.

OPL Model and Input Data

OPL Model: Type or paste the Optimization Programming Language (OPL) model syntax in this field.

Tuple Set Name in OPL Corresponds to Incoming Data: Enter the tuple set name in the OPL model that corresponds to the incoming data. Then, if needed, verify that all tuple fields are mapped to data input fields according to their order in the tuple definition.

Input Mapping: Enter the tuple fields and data input fields for mapping. All tuple fields must be mapped to data input fields in the order they're declared in the tuple definition.

Other Data

OPL Data: Use the Other Data tab if you need to specify any other data for the optimization.

Output

When the output is a decision variable, it must take the prior data sources (incoming data) as indexes, and the indexes must be predefined in the Input Mappings section on the OPL Model and Input Data tab. No other type of decision variables are currently supported. The decision variable can have a single index or multiple indexes. SPSS Modeler will output the CPLEX results with all or part of the original incoming data together, which is consistent with other SPSS Modeler nodes. The referred corresponding indexes must be specified in the Output Tuple field described below.

Output Mode: Choose the output mode (Raw Output or Decision Variable) and specify other options as appropriate. The Raw Output option will output the objective function value directly, regardless of name.

Objective Function Value Variable Name in OPL: This field is enabled if you selected the Decision Variable output mode. Enter the name of the objective function value variable from the OPL model.

Objective Function Value Field Name for Output: Enter the field name to use in the output. The default is _OBJECTIVE.

Output Tuple: Enter the name of the predefined tuple from the incoming data. This acts as the indexes of the decision variable and is expected to be output with the Variable Outputs. The Output Tuple should be consistent with the decision variable definition in the OPL. If there are multiple indexes, the tuple names must be joined by a comma (,).

Variable Outputs: Add one or more variables to include in the output.

Note:

When running a flow containing a CPLEX Optimization node, the CPLEX library has a limitation of 1000 variables and 1000 constraints.