Content Edited in protest for the bad direction of the Home-Asisstant project.
Bloody hell, that info is far too scary. I probably don’t want to know!
2 Likes
I understand the concern. However, quantified-self is not a bad thing as long as it’s private and useful.
If you meant you don’t want to know other people numbers, that’s fair and I change the numbers in the HTML before sharing 
No I mean I don’t want to know about my own data.
1 Like
I’d like the idea that can track the calorie counter as well. From a dietary POV. But definitely from its own database and collecting the calorie info from the USDA.
Most of which is incredibly unreliable…
Just wanted to say thanks for this incredibly detailed post. This has inspired me to start doing something similar and I imagine there are more than a few New Year’s resolutions that could do with being tracked in a similar fashion.
1 Like
Happy to hear. I would love to hear what you will build (or have built when done), if you are willing to share 
QS is something I’m very interested in. That said, I think HS is probably the wrong place to build such a dashboard? I have been using the predecessor to Heedy (https://github.com/heedy/heedy/) to store and visualize some data. And I’m currently playing with Graphite/Grafana as a way to do that as well. I’m pushing pollen & mold data into it so I can try to see how that’s affecting me.
My goal is to fix sleep problems, so I have an Oura. What I’m trying to figure out next is how to conveniently collect medication, exercise, and food data. Is Allegra screwing up my sleep while Clairitin is fine? If I try l-theanine or melatonin, do they help deep sleep? It’s really that collection side of things that is giving me fits.
1 Like
Hey @nitobuendia, thanks for the response.
For awhile, I used Tickmate to feed into ConnectorDB (the previous name for Heedy). It’s a nice app. I wrote some python that would take the Tickmate data dump and push it into ConnectorDB https://github.com/robbieh/tickmate-connectordb
This was kind-of great, but I found it wasn’t flexible enough in the long run. I was able to do some of the analysis directly in ConnectorDB, but ended up dropping into R to explore more. I like that Heedy is including Jupyter notebook capability. That would make doing the analysis easier. I have used Jupyter in other contexts, but haven’t got Heedy up and running yet.
I still haven’t figured out the data collection for things like medication doses, though. It’s unfortunate InfluxDB doesn’t do joins. I had just started to explore using it as a data store.
1 Like