Assess PML scalability

[kevin-delmas]

Provide a set of synthetic benchmark of models to assess PML analyzer scalability

[kevin-delmas]

I have updated the code to enable extra computation options to disable semantics estimation and multi-transactions exports for experimentations. In all cases, a summary contains the number of free and itf per size and computation time. This would avoid to dump irrelevant information.

I will try to provide a test to compute complexity and reduction metrics.

@Gaudeval we should discuss what kind of data visualisation would be relevant here. There is a draft of GNUPlot exporter for figures. Perhaps to be discussed in a new issue related to data visualization