Contraintuitive index handling of simulation time

[kim-mskw]

Currently, the end of the simulation is translated confusingly to the index, which used to test the correct length of the data frame.

At the end of the simulation, 00:00 seems to belong to the next day, while at the start, it belongs to the former. The end is 03.08 00:00, which lets the simulation run till 03.09 00:00 to compare the index.

Sometimes, in the examples, we artificially extend the index by 24 hours. Compare tutorial 03 in the handling of the electrolyzer. I guess this is because we want to fit the weird translation of the end of the simulation, but for beginners, it seems simply confusing.

[kim-mskw]

Is there any reason behind that?

[nick-harder]

@kim-mskw I believe we simply add extra 24 hours to the index so the forecasting things and market clearings work properly. But I also don't like this solution and we should remove this confusing thing

[nick-harder]

@maurerle when looking into the code and removing the extra one day we had, I don't see any issues with the code. The history says you added that one extra day. Do you remember why exactly?

index = pd.date_range( start=start, end=end + timedelta(days=1), freq=config["time_step"], )

[maurerle]

This is correct:

I believe we simply add extra 24 hours to the index so the forecasting things and market clearings work properly

I think if you always bid for the next 24 hours at once, we had an issue that the forecast was accessing values after the end of the simulation. I am not 100% sure if this is still needed, as we improved ending the simulation at some place. Though I thought it was not needed anymore last time and was proofed wrong :laughing: