Why are some environmental conditions arbitrarily discretised?

[mansoor2016]

The precipitation intensity is provided as a discrete enum. why? It is not consistent for all env. conditions.

If you develop regression models for environmental effects, likely you will need continuous inputs.

Providing a double value field + a confidence level field would be a more useable design.

[jdsika]

Hey!
If you look at the PRs you can find some discussions around the topic and also one with your suggestion: https://github.com/OpenSimulationInterface/open-simulation-interface/pull/582/files

Environmental conditions are often not measured very precisely, imagine a real world drive somewhere and the driver would need to document and classify the current rain intensity. Having a double field would just create the illusion of precision. If your modlling requires discrete values I recommend to make a proposal to add e.g additional continous values without breaking backwards compatibility and allowing for more precise values not allowing them to contradict.

new_value = 0.3
precipitation = PRECIPITATION_VERY_LIGHT

// Very light intensity of precipitation, when I in [0.1,0.5[ mm/h
//
PRECIPITATION_VERY_LIGHT = 3;

There is also a modelling working group and I suppose they have formed some opinions around it alrady. @PhRosenberger may know something here?

Best regards
Carlo

[PhRosenberger]

Yes, I have already collected several discussions/issues/PRs on that topic and I would be glad if you, @mansoor2016, would join our project to elaborate on that topic!

Here is a starting point for that: #601

[mansoor2016]

Looks like the suggestions in #601 would resolve this discussion, closing this now.

Thanks @jdsika & @PhRosenberger