Time series versions

[borlaugen]

Is it possible to read multiple versions or a specific version of a time series using the Python SDK?

[tnoczyns-volue]

What do you mean by specific version of time series?

• Historical version: https://volue-public.github.io/energy-mesh-python/mesh_functions.html#history
• Specific vl code
• Something else?

[borlaugen]

This looks like what I want. Thanks!

Is there an example of e.g. getting all versions of a forecast given by a target object?

[tnoczyns-volue]

Not sure if this is what you need, but for getting historical versions:

# Get historical time series
# Calculation expression: ## = @GetTsHistoricalVersions(@t('.some_ts_attribute'),5)

LOCAL_TIME_ZONE = tz.tzlocal()
start = datetime(2021, 9, 1, tzinfo=LOCAL_TIME_ZONE)
end = datetime(2021, 9, 15, tzinfo=LOCAL_TIME_ZONE)

max_number_of_versions_to_get = 5
historical_timeseries = session.history_functions(
    timeseries_attribute.id, start_time=start, end_time=end
).get_ts_historical_versions(max_number_of_versions_to_get)

for timeseries in enumerate(historical_timeseries):
    pandas_dataframe = timeseries.arrow_table.to_pandas()
    # Post processing: convert to UTC timezone-aware datetime object and then to given time zone
    pandas_dataframe["utc_time"] = pd.to_datetime(
        pandas_dataframe["utc_time"], utc=True
    ).dt.tz_convert(LOCAL_TIME_ZONE)

For getting all forecasts:

# Get forecast time series
# Calculation expression: ## = @GetAllForecasts(@t('.some_ts_attribute'))

LOCAL_TIME_ZONE = tz.tzlocal()
start = datetime(2021, 9, 1, tzinfo=LOCAL_TIME_ZONE)
end = datetime(2021, 9, 15, tzinfo=LOCAL_TIME_ZONE)

forecast_timeseries = session.forecast_functions(
    timeseries_attribute.id, start_time=start, end_time=end
).get_all_forecasts()

for timeseries in enumerate(forecast_timeseries):
    pandas_dataframe = timeseries.arrow_table.to_pandas()
    # Post processing: convert to UTC timezone-aware datetime object and then to given time zone
    pandas_dataframe["utc_time"] = pd.to_datetime(
        pandas_dataframe["utc_time"], utc=True
    ).dt.tz_convert(LOCAL_TIME_ZONE)

Similar for other Mesh functions: https://volue-public.github.io/energy-mesh-python/mesh_functions.html.

[borlaugen]

Thank you, this helped a lot. Is there a way to distinguish from the contents of a volue.mesh._attribute.TimeseriesAttribute whether or not it has forecasts?

[tnoczyns-volue]

Is there a way to distinguish from the contents of a volue.mesh._attribute.TimeseriesAttribute whether or not it has forecasts?

Unfortunately no, at least none that I know of.
A time series attribute without forecasts won't return any time series when called with get_all_forecasts.