energy_analysis_toolbox.load_profiles.thresholds.relative_std module#

Module for computing threshold profiles based on relative standard deviations.

This module provides tools to calculate profiles that deviate from the mean profile by a user-defined multiple of the standard deviation derived from historical data. The profiles aim to capture variability while maintaining a consistent baseline.

class energy_analysis_toolbox.timeseries.profiles.thresholds.relative_std.RelativeSTDThreshold(offset_std: float = 3, **kwargs)[source]#

Bases: MeanProfile

A class which implements a statistical deviation from the mean profile.

compute(history: Series, time: Timestamp, **kwargs) → Series[source]#

Return a threshold profile.

The threshold profile is obtained using a user-defined relative variation from the mean profile built from history.

Parameters:

  • history (pd.Series) – Consumption history used to computed the reference profile.

  • time (pd.Timestamp) – The time at which the profile is of interest. Only the information about the date is used in the passed timestamp.

Returns:

profile (pd.Series) – Profile threshold with same resolution as the history data

Notes

The profile threshold is obtained as the mean profile obtained from history + tshd times the standard deviation profile.