Home Data

Feel free to reach out to me on Twitter to discuss: martindaniel4.

Background

In July 2018 my wife and I bought and revamp a home in Paris, France. We love our place but one issue we have is the level of heat we experience. Since our flat has a single-glass ceiling (or "Verrière") the sun hits pretty hard and the prospect of extreme heat waves happening in Paris is worrysome.

In February 2019, I acquired a Netatmo thermostat (see - https://www.netatmo.com/fr-fr/energy/thermostat) that I connected to my heating system. In addition to remote control of my home temperature, it also collects temperature. Since Netatmo has an API I figured I would give it a try.

A picture of my Netatmo thermostat in my apartment

Here is for instance a first graph of those temperature, since I have installed my thermostat:

Interior temperature graph in my home since Feb. 2019 (step = 30min, celsius)

From there I have also started collecting data from other devices such as Gas (GAZPAR / GRDF) or Electricity (Linky / ENEDIS).

Usage

PYTHONPATH

You first want to add the path of that repo to your PYTHONPATH so it is easily accessible when running a script. You can do so in your bash profile with:

export PYTHONPATH="/path/to/compans_data:$PYTHONPATH"

I also store my credentials in my bash profile. Those can then be accessed from a Python script with the os library:

import os
email = os.environ["email"]

Temperature

Netatmo

The logic is coded in Python in netatmo_temperature.py.

• First create an app on Netatmo platform (see https://dev.netatmo.com/)
• Then add to your path the following variables: client_id, client_secret, email and password.
• Install Python libraries with pip install -r requirements.txt
• You can now retrieve all your temperature with the following command:

from netatmo_temperature import * 
pull_temperature()

Exterior

I was not able to easily retrieve a temperature dataset over the last 2 years in Paris with a subday step. I found that surprising. On the advice of ssaunier I ended up scrapping Meteo Ciel to retrieve an hour step measure of Paris temperature. The scrapping logic can be found in temperature_scraping.py and the csv result in temp_all_paris.csv.

Gas

The logic is stolen from empierre and is available in the gaz folder.

• Make sure you have a GAZPAR account.
• Add your GAZPAR_EMAIL and GAZPAR_PWD credentials in your bash profile.
• You can retrieve your gaz consumption with:

from gaz.gazpar import get_data_per_day
get_data_per_day('30-07-2020', '07-08-2020')

Next Steps

My main goal is to retrieve measures of interior temperature on a long timeframe. My hope is that I can start running some analysis and inform some home decisions as I revamp my home. For instance:

• What's the impact of covering my ceiling from the interior?
• What's the impact of sun or cloud on interior temperature?
• How much temperature do I save if I change for a double glass-ceiling?

Building on that, I also want to incorporate additional measures such as electricity consumption, hygrometry and more!