detect_disruptive_period: Detect disruptive periods in consumption time series.

In biggproject/biggr: biggr

Detect disruptive periods in consumption time series.

Description

This function detects a disruptive period in a consumption time series. It evaluates different date ranges to find the most suitable one containing a disruptive period that should not be considered in the training phases of statistical models. An example of a disruptive period could be the Covid lockdown when building consumption radically changes due to the different occupancy and activity patterns. The model used to detect the disruptive period considers the consumption relationship between weekdays and temperature.

Usage

detect_disruptive_period(
  data,
  consumptionColumn,
  timeColumn,
  temperatureColumn,
  tz,
  minIniDate,
  maxIniDate,
  minEndDate,
  maxEndDate,
  checkFor = "decrement",
  minDecrementPercentualAffectation = 30,
  minIncrementPercentualAffectation = 60
)

Arguments

data	<data.frame> Time series with potential anomalies in values. It should contain a time column, a consumption column and a temperature column.
consumptionColumn	<string> Consumption column in data time series
timeColumn	<string> Time column in data time series
temperatureColumn	<string> Temperature column in data time series
tz	<string> specifying the local time zone related to the building in analysis. The format of this time zones are defined by the IANA Time Zone Database (https://www.iana.org/time-zones).
minIniDate	<date> Minimum date to start looking for anomalies
maxIniDate	<date> Maximum date to start looking for anomalies
minEndDate	<date> Minimum date to stop looking for anomalies
maxEndDate	<date> Maximum date to stop looking for anomalies
checkFor	<string> Which effect will the disruptive event generate? e.g. the SARS-CoV2 pandemia decrease the energy consumption of tertiary buildings, so in that case we should check for decrement in consumption Possible values: decrement, increment, incrementAndDecrement.
minDecrementPercentualAffectation	<float> indicating the minimum decrement of consumption (vs. baseline) to be considered as abnormal. Default 30%.
minIncrementPercentualAffectation	<float> indicating the minimum increment of consumption (vs. baseline) to be considered as abnormal. Default 60%.

Value

<data.frame> with the period of time (min-max) with abnormal consumption.

biggproject/biggr documentation built on Oct. 2, 2024, 11:13 p.m.