Data mining and jupyter

I like to extract useful data from sqlite and convert it to useful information through analysis.

https://www.pacienciadigital.com/analisis-de-consumo-de-energia-con-home-assistant/

Do you have more cool ideas of analysis with jupyter? Some correlation between sensors, classification or regression with machine learning algorithms? Random forest? Ideas are welcome.

1 Like

Ah so sad that this is only in spanish, wanted to take a closer look at this because i wanted to do some timeline analysis for the fuel prices to predict best tanking time daily

Do you know the Home Assistant Data Science Portal? : https://data.home-assistant.io/

Maybe that’s a good starting point. There where also some usage examples back in time but couldn’t find them anymoree. What happend to them @robmarkcole ?

1 Like

There are also some example Notebooks in the official Home Assistant Repo: home-assistant-notebooks/other at master · home-assistant/home-assistant-notebooks · GitHub

1 Like

I’m fact I was thinking to translate it, as I can see more interest on that topic among the English-spaeking folks.
I’m sure it will not be a 100% correct, but good enough to read it with no problem.
I’ll let you know.

1 Like

Hi Simon. Yes I saw some information on HA data science portal. That’s exactly the idea. I can see we are not using our own rich data from HA.
I’m eager to find some applications (real applications).
My raspberry is not powered enough, but I can always extract the data, find some patterns, or fit some simple rules and go ahead with some information and build in an automation in my system.

Other examples with bayesian sensor:

https://www.pacienciadigital.com/sensor-bayesian-estadistica-para-tu-sistema-domotico/

Or just using DB access to optimize:

https://www.pacienciadigital.com/como-optimizar-la-base-de-datos-de-home-assistant/

I ended up with an English version:

https://xxx.pacienciadigital.com/energy-data/

Ideas on how to use database data are welcome.

Regards.