Format of time series

[SabineHaas]

It should be possible to enter time series with a datetime column.

• [ ] check if this is possible now - if not implement
• [ ] check if conversions/calculations are done correct (@smartie2076 you had doubts)

[smartie2076]

* [ ]  check if this is possible now - if not implement

Well - for files that are defined for parameters that could also be scalars (via "{'value': {'file_name': 'Walqa_data.csv', 'header': 'consumption_price'}, 'unit': 'currency/kWh'}"), a datetimestamp is possible. I would like the feature enable this for energyConsumption (see issue #179) and energyProduction.

* [ ]  check if conversions/calculations are done correct (@smartie2076 you had doubts)

Does this relate to #195? My doubts related to time steps - we currently only use hourly steps, not another frequency.

[SabineHaas]

Does this relate to #195?

Yes, from my point of view we can close #195 and discuss here, as tests should be part of this issue anyways. `

[Bachibouzouk]

This is related to #626, from the EPA we take their input file without the datetime column, we assume all series match perfectly the number of time steps starting from start_date over evaluated_period.

We can deal with #196 by modifying receive_timeseries_from_csv to also accept a datetime column.

One think that the person implementing this should pay attention to is:

• if the loaded datetime index does not match the datetime index build from start_date, time_steps and evaluated_period what should be done? (i.e. warning, error, ??) The best would be to write tests for all possible combinaisons (start_date and evaluated_period match but not time_steps or start_date and time_steps match but not 'evaluated_period', etc.)