End-assignment-only results in confusing convergence message

[zptro]

I realized that this line which was added in #537 makes the try clause succeed, even if self.old_car_demand is missing, which in the case of end-assignment-only runs results in a confusing message that the model run has not converged. The intention is originally that the try clause should fail in the first iteration, which forces relative gap to zero.

helmet-model-system/Scripts/assignment/departure_time.py

Line 54 in dc32a3e

self.old_car_demand = cast(numpy.ndarray, self.old_car_demand) #type checker hint

[hsl-petrhaj]

@zptro
I investigated this a bit:

1. End-assignment-only always sets the iterations in helmet.py to zero, thus the warning in roughly helmet.py:125 always gets triggered.
2. Cast should not do anything in runtime, but since I do not know all its intricacies it is possible that for example None would get turned into something else.
3. Is it possible that old_car_demand is something else than int or ndarray? Then we should also adjust the type annotation.