energy_forecast.consumption_forecast module#

class energy_forecast.consumption_forecast.DayForecast[source]#

Bases: TypedDict

end_date: str#
peak: PeakValue#
start_date: str#
updated_date: str#
class energy_forecast.consumption_forecast.OneValue[source]#

Bases: TypedDict

end_date: str#
start_date: str#
value: int#
class energy_forecast.consumption_forecast.PeakValue[source]#

Bases: TypedDict

peak_hour: str#
temperature: float#
temperature_deviation: float#
value: int#
class energy_forecast.consumption_forecast.PredictionForecast[source]#

Bases: TypedDict

weekly_forecasts: list[DayForecast]#
class energy_forecast.consumption_forecast.PredictionForecastAPI(secret: str | None = None)[source]#

Bases: RTEAPROAuth2

Access the RTE API to get the weekly forecast of consumption.

format_short_term_data(json_data: dict) → DataFrame[source]#
format_weekly_data(json_data: PredictionForecast) → DataFrame[source]#

Format the raw data from the API into a DataFrame.

Parameters:

json_data (PredictionForecast) – The raw data from the API.

Returns:

pd.DataFrame – The formatted data. The Index is the date of the prediction. Includes the columns:

  • predicted_consumption: the predicted consumption in MW

  • predicted_at: the date when the prediction was calculated

get_short_term_json(start_date: str | Timestamp | None = None, end_date: str | Timestamp | None = None, horizon='3d')[source]#
get_weekly_forecast(start_date, end_date=None, horizon='1w')[source]#

Retrieve the weekly forecast of consumption.

Parameters:

  • start_date (str, Timestamp) – the start date of the forecast.

  • end_date (str, Timestamp, optional) – the end of the forecast, by default None If None, the forecast is for the next horizon.

  • horizon (str, Timedelta, optional) – If end_date is none, the duration of the forecast from start_date , by default “1w”

Returns:

pd.DataFrame – The weekly forecast of consumption.

See also

-

py:meth:get_raw_data

-

py:meth:format_weekly_data

get_weekly_json(start_date: str | Timestamp | None = None, end_date: str | Timestamp | None = None, horizon='1w') → PredictionForecast[source]#

Retrieve the raw data from the API.

Parameters:

  • start_date (str, Timestamp) – the start date of the forecast.

  • end_date (str, Timestamp, optional) – the end of the forecast, by default None If None, the forecast is for the next horizon.

  • horizon (str, Timedelta, optional) – If end_date is none, the duration of the forecast from start_date , by default “1w”

Returns:

PredictionForecast – The raw data from the API.

url_api_short = 'https://digital.iservices.rte-france.com/open_api/consumption/v1/short_term'#
url_api_weekly = 'https://digital.iservices.rte-france.com/open_api/consumption/v1/weekly_forecasts'#