What are some good tools for graphing sensor values?

Question

I'm using Home Assistant for my home automation, it has some basic graphing capabilities but only for the last 24 hours. I'd like to create graphs both with fine granularity and over long periods of time.

I have a home office in the basement, where I smoke cigars. I have a large extraction fan with a pretty serious airflow that pulls the smoke out and keeps the room in negative pressure so the smoke and fumes doesn't reach any other parts of the house.

I am interested in seeing how the different fan speeds affects the temperature in other rooms in the basement, and how it again is affected by the outside temperature so e.g. summer vs winter.

For this I need to plot various temperature and other sensor data over long periods of time, and I need to be able to see how the different sensor readings correlate (or not).

What are some good graphing tools that are easy to interface with home automation and IoT sensors or the Home Assistant?

Answer 1

InfluxDB in Domoticz

This is a database provided with the Domoticz installation on Raspberry Pi. It allows to use data with time series.

Domoticz uses this to modelize sensors' data

Answer 2

For Home Assistant specifically, you can connect to the SQLite database and use your own graphing software (or a script) to generate custom graphs. The Home Assistant blog demonstrates the use of Python with matplotlib to do this:

# Adapted from the linked code from Home Assistant.
import sqlite3
from matplotlib import dates
import matplotlib.pyplot as plt

import homeassistant.util.dt as dt

ENTITY_ID = 'entity id here'
START_DATE = 'date here'
END_DATE = 'date here'

values = []
timestamps = []

conn = sqlite3.connect('/home/ha/.homeassistant/home-assistant_v2.db')
data = conn.execute("SELECT state, last_changed FROM states WHERE entity_id = {} AND last_changed BETWEEN {} AND {}".format(ENTITY_ID, START_DATE, END_DATE))

for x in data:
    timestamps.append(dates.date2num(dt.parse_datetime(x[1])))
    values.append(float(x[0]))

plt.plot_date(x=timestamps, y=values, fmt="r-")
plt.ylabel('Value')
plt.xlabel('Time line')

plt.savefig('sensor.png')

The database schema is available here. What we are interested in is state objects; you should know the entity_id of the device you're interested in.

If you're familiar with Python, adapting that should be relatively easy, and you could even add a GUI or nicer command-line interface. Any language that can query the SQLite database will work fine, though.

Alternatively, you might consider exporting to CSV and using a spreadsheet program—no doubt this will be more challenging to automate, but may be more user-friendly if you're not a programmer.