citylearn.internal.kpi module

class citylearn.internal.kpi.CityLearnKPIService(env: CityLearnEnv)[source]

Bases: object

Internal KPI/evaluation service for CityLearnEnv.

EV_DEPARTURE_WITHIN_TOLERANCE_DEFAULT = 0.05
LEGACY_COST_FUNCTIONS = {'all_time_peak_average', 'annual_normalized_unserved_energy_total', 'carbon_emissions_total', 'cost_total', 'daily_one_minus_load_factor_average', 'daily_peak_average', 'discomfort_cold_delta_average', 'discomfort_cold_delta_maximum', 'discomfort_cold_delta_minimum', 'discomfort_cold_proportion', 'discomfort_hot_delta_average', 'discomfort_hot_delta_maximum', 'discomfort_hot_delta_minimum', 'discomfort_hot_proportion', 'discomfort_proportion', 'electricity_consumption_total', 'monthly_one_minus_load_factor_average', 'one_minus_thermal_resilience_proportion', 'power_outage_normalized_unserved_energy_total', 'ramping_average', 'zero_net_energy'}
evaluate(control_condition=None, baseline_condition=None, comfort_band: float = None, *, evaluation_condition_cls, dynamics_building_cls) → DataFrame[source]

Evaluate cost functions at current time step.

evaluate_legacy(control_condition=None, baseline_condition=None, comfort_band: float = None, *, evaluation_condition_cls, dynamics_building_cls) → DataFrame[source]
evaluate_v2(control_condition=None, baseline_condition=None, comfort_band: float = None, *, evaluation_condition_cls, dynamics_building_cls) → DataFrame[source]