This is the basic building block in intuceo. Developer has to pass sub analysis id for all subsequent analyses steps. If the analysis on hand requires minor changes in the way analysis needs to be carried out (eg. Choosing which attributes are important), a separate sub analysis can be created. All sub analyses share the same target.
Intuceo analyses are organized as below: Dataset –> Analysis –> Sub-Analysis. For each dataset, one can have many analyses. An analysis is defined with its name and the target attribute for which analysis is to be performed. Multiple sub-analyses can be performed to fine tune/play around various parameters to arrive at final sub-analysis. However, all these sub-analyses have the same target attribute of the analysis under which this sub-analysis is created. If user wants to do another analysis on same dataset for different attribute as target in the dataset, user needs to create new analysis. So, hierarchy of analyses are multiple datasets –> multiple analyses; each has its own target attribute –> multiple sub-analyses under each analysis. While this is a bit of extra step of sub-analysis creation, user will get benefit of organized analysis space than a space with bunch of analyses.