Multiple Clustering (5.0.4)
Upon completion of the Multiple Clustering process, the generated Factors can be automatically connected to their manifest variables.
As a result, one obtains a set of naive structures, which can subsequently be used for Probabilistic Structural Equation modeling (PSEM).
Most of the algorithms that are based on random number generation come now with the possibility to fix and define the Seed.
It is then possible to reproduce the results.
Possible PSEM Workflow
- Select and fix all the arcs (Ctrl+A - Right Click on one arc - Properties - Fix)
- Include the Target node
- Define forbidden arcs to prevent any additional connections with the Manifest variables
- Perform Taboo Learning for keeping the arcs that have been fixed
The Fixed arcs are now shown with dotted arcs (and not as blue arcs anymore)
