Awesome

README

This is a Raspberry Pi temperature graphing PoC through Plotly (usually up to 200 points drawn) and persistant storage through Google Docs (Spreadsheet).

Plots CPU temperature (directly from RPI), environment temperature (BMP085), environment barometric pressure (BMP085), environment humidity (DHT22). and outdoor temperature (Weather Underground). It runs as a Unix daemon and preferably runs infinitely long.

If there is a LED available, it will pulse it in the background to indicate its running status.

Snapshot

/rpi-plot.png /rpi-board.png

Hardware requirements

Raspberry Pi model A or B: http://shop.pimoroni.com/
dupont cables or cobbler with breadboard (GPIO and I2C): http://www.smart-elex.co.uk/RaspberryPi/RASPBERRY-Pi-GPIO-ACCESSORIES/RPI-Electronics-kit1
BMP085, BMP180, BMP183: https://www.adafruit.com/products/1900
DHT11, DHT22 or DHT2302: https://www.adafruit.com/products/385
LED
10kOhm and 330Ohm resistors (for DHT and LED)

Software Requirements

Weather Underground developer account (free!): http://www.wunderground.com/weather/api
Plotly account (free!): http://plot.ly
Google Docs account (optional)
WiringPI library: git://git.drogon.net/wiringPi
bcm2835 library: http://www.airspayce.com/mikem/bcm2835/
Adafruit BMP085 I2C library: included!
Adafruit DHT GPIO library: https://github.com/adafruit/Adafruit_Python_DHT
Plotly library: pip install plotly
daemon library: pip install daemon
gspread library: pip install gspread

Important notes

Raspberry PI model A users need to edit Adafruit_I2C.py and do the following change:

    self.bus = smbus.SMBus(0);

You can store Weather Underground configuration in /root/.weather_underground.rc:

    {"wu_city": "Zagreb", "wu_state": "Croatia", "wu_key": "XXXX"}

Plotly configuration needs to be stored in /root/.plotly/.credentials
You can store Google Docs configuration in /root/.google_docs.rc:

    {"gdocs_email": "somebody@gmail.com", "gdocs_password": "secret password", "gdocs_sheet": "somesheet"}

Monitoring

Integration with Supervisor http://supervisord.org/ process control system is rather trivial. What is needed is a corresponding daemon configuration in /etc/supervisor/conf.d/rpi-plot.conf if the code has been placed at /home/pi/work/rpi-home-sensors:

    [program:rpi-plot]
    command=/home/pi/work/rpi-home-sensors/rpi-plot.py nodaemon
    stopsignal=INT
    autostart=true
    autorestart=true
    redirect_stderr=true
    stdout_logfile=/var/log/rpi-plot.log
    environment=HOME="/root"

Then the simple sequence of the following commands is enough to kickstart the process:

    supervisorctl update
    supervisorctl status