You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Inconsistent time index handling in BDEW profiles #104
The time index is handled inconsistently across the BDEW profile classes, and the docs don't say what's expected.
Outdated docstring in HeatBuilding HeatBuilding.__init__(self, df_index, **kwargs) builds its DataFrame from df_index, but the docstring only documents a year parameter, which the code doesn't use.
Timezone of the index is not documented ElecSlp(year) creates a timezone-naive index internally, and the example in get_scaled_power_profiles uses naive timestamps. For HeatBuilding, the docs don't mention timezones at all. However, get_sf_values() and get_weekday_parameters() rebuild naive indices from date + hour, which suggests df_index should be timezone-naive.
Proposed changes
Replace year with df_index in the HeatBuilding docstring.
State explicitly in the docs that the index must be timezone-naive (if that is the intended behaviour).
Feature request: df_index for ElecSlp
Current behaviour:ElecSlp(year) always returns one full calendar year at 15-min resolution. For model periods that don't start on Jan 1, users have to cut and append yearly profiles manually.
Requested behaviour: Allow ElecSlp to accept a df_index (any start, length and resolution), analogous to HeatBuilding. The year parameter could remain for backwards compatibility.
Activity