Model booleanization
Description
Model booleanization consists in mapping a multivalued model on Boolean variables.
bioLQM uses the mapping originaly proposed by van Ham, in which a component associated with the maximal value m will be mapped on m Boolean components. For example, a component taking the values 0, 1, 2, and 3 will be encoded as 000, 100, 110, and 111.
Handling non-admissible states
The booleanization introduces many non-admissible states, which may require special care depending on the analysis applied on the booleanized model. This modifier makes sure that a simulation which start with an admissible state will not explore non-admissible states. It also prevents the introduction of non-admissible attractors by making sure that at least one admissible state is reachable from any non-admissible.
Usage
The booleanization is a simple modifier without parameters: just use the -m booleanize switch on the command line or lqm.modifyModel(model, "booleanize") in scripts.
Note that booleanization can also be used automatically to export multivalued models in formats supporting only Boolean components.
Going further
This service is provided by the BooleanizeService.