The data provided consists of the reading in the main meter and two sub-meters of five buildings from 01-04-2017 to 31-12-2017. The given timestamp is divided into 15-minute intervals. The dataset consists of electricity consumption of five buildings across three different meters for each timestamp.
The table and graph below describes the statistical property of dataset.
| Main Meter | Sub Meter 1 | Sub Meter 2 | |
|---|---|---|---|
| Mean | 6095.76 | 2092.88 | 776.83 |
| Std | 3485.12 | 1135.88 | 883.49 |
| Min | 61.63 | 0.00 | 0.00 |
| 25% | 3455.91 | 1365.00 | 4.85 |
| 50% | 5362.18 | 2075.67 | 621.91 |
| 75% | 7685.40 | 2772.50 | 1365.89 |
Figure 1: Graph of all 15 timeseries
Train.csv have raw dataset in the interval of 15 minutes. In Intermediate files, the CSV files have dataset in interval of 1 hours.
Three methods are implemented.