Pointers

The structure of a Dakota study is specified using string-based pointers and block IDs. Except for environment, of which there can be at most one per study, each block (e.g. method, model) can have a unique ID, which is provided using the id_<block> keyword. This makes the block referenceable by other blocks, typically using <block>_pointer keywords.

For simple studies, where there is only one of each block type in the input file, use of IDs and pointers to disambiguate structure is not required. However, they may still be used to clarify output.

Example

Fig. 61 illustrates how pointers and IDs are used to specify the structure of a Dakota study using colored arrows.

Dakota input with IDs and pointers that features arrows to indicate connections

Fig. 61 The colored errors draw attention to pointers and IDs that define the structure of the study