Desired QA values are assumed lower than NFR and in [0.0, 1.0]

[chcorbato]

Currently, diagnosis of quality attributes assumes that

• QA values are in [1.0, 0.0],
• the desired QA values need to stay lower than the NFR value

The convention in this repo is different in the repo about predictive observers by https://github.com/JWijkhuizen

• [link to repo needed here]

Is this the best modelling of QAs? Can we model it better by defining a "metamodel" of QAs in TOMASys and leaving the definition of the QA metric to the domain?
Should we make it explicit? Can we move this model to the ontology (out of the python code)?

mc_mros_reasoner/mros1_reasoner/src/mros1_reasoner/tomasys.py

Line 211 in 0f1b05d

if qas[0].hasValue < nfr.hasValue: