energy_anomaly_detect: anomaly_detect_package

In jaeseokjeong/detect_sensor_outlier:

Description

This package judges whether or not the power consumption per hour is abnomaly.

Usage

energy_anomaly_detect(selected_data, weather_data, training_start_date,
  training_finish_date, test_start_date, test_finish_date)

Arguments

selected_data	The data variable should be composed of the household number (id), date (date), and power usage (usage).
weather_data	It is the data extracted by Korean weather (koreaweather) R package. The variables in the data are date, average temperature, relative humidity, and wind.
training_start_date	Determine the start date of training data to use in setting the prediction model in the R package.
training_finish_date	Determine the end date of the training data to use in setting the prediction model in the R package.
test_start_date	Set the start date of test data to detect anomaly.
test_finish_date	Set the finsih date of test data to detect anomaly.

Examples

 energy_anomaly_detect(example_household, weather_information,
 "2016-01-03 23:00:00","2016-01-21 23:00:00",
 "2016-01-22 00:00:00", "2016-01-22 05:00:00")

jaeseokjeong/detect_sensor_outlier documentation built on May 24, 2019, 9:54 a.m.