The data was provided by the LERTA company (https://www.lerta.energy/). It was collected as part of a project aimed at improving methods of controlling thermal comfort in buildings. The data refers to an office building located in Poznan and it contains:
- room temperature (for several temperature sensors located in different places) [°C],
- the degree of opening of the radiator valve [%],
- set temperature [°C],
- the size of the room,
- the number of people in room,
- the window orientation and placement.
#The aim of the project is to make two predictions:
- average temperature value of the indicated sensor for 8 hours (from 6:00 a.m. to 2:00 p.m.),
- the value of the average valve opening degree in 8 hours (from 6:00 - 14:00)
Project is created with:
- Python version: 3.8.6
- Pandas version: 1.1.5
All the necessary libraries are in the requirements file. You can install with a command:
$ pip install -r requirements.txt
To run this project, you need to pass two arguments:
- path to input json file,
- path to output result csv file
$ python main.py input_file results_file_csv
You can find both models in file models. There are files that create models.p used in program. To predict temperature linear model Ridge was used. To predict valve RandomForestRegressor was used.
To evaluate results Mean Absolute Error was used.