Should coralslave write to .csv when in initialization phase?

[mrindar]

During the initialization phase (is that the correct term?) when coral does the initial fmiDoStep() calls with step_size=0, coralslave still writes the values to the output .csv file - this looks strange when plotting. One way to get around this is to post process the .csv file of course, but that post processing then depends on how Coral does its initialization phase (currently there is one fmiDoStep() for each slave in the simulation) which I don't like. I think the best solution is to simply have coralslave not write to .csv during this initialization phase

[kyllingstad]

Personally, I've found this to be useful for debugging, to see how initial values propagate and influence each other. But I definitely see your point, and I'll look into excluding those steps from the output.