With the help of Homey’s insightful features, you can view all sensor measurements and their historical evolution. Take it a step further and do more than just observe the insights; act accordingly or even glimpse the future a little.
You can choose an insight (a device and a capacity, such as the outside temperature), as well as a timeframe to analyze. The outcomes will then be computed as follows:
a number between -1 and 1 that indicates whether the sensors’ measurements are increasing or decreasing (and by how much/slowly). For instance, a score of 0.75 indicates a sharp upward trend, while -0.1 denotes a mild negative trend.
the smallest/largest value found within the timeframe
the average value across all data points (arithmetic mean)
If all values were sorted and you took the middle value, it would be the median.
Passing in 50% as an argument will return the median, while 25% and 75% will return the first and third quartiles respectively.
a measure of how “spread out” the data set is.
Number of used data points
The number of measures in the dataset. This can help assess how reliable the results presented above are. Important: A larger timeframe may not contain a larger dataset than a smaller one. Since Athom decreases the data points over a longer timespan.
Feedback / Bugs
Any requests should be made in this thread or in the issues area on Github
Wanna buy me a coffee or help to get the new Homey pro?
Feel free to donate to me for my work It’s highly appreciated!
Check out my other apps
(This app idea is based on the app from MadMonkey87, but rewritten and republished by me)
Reserved for future stuff.
The app just got released!
Just out of curiosity, is it compatible with previous version or it behaves as fully new app so it is required to reset everything ? Regardless of answer, thank you very very much for revival
update OK, fully new app, flows need to be reworked*
** update 2 ** Do I miss something ?
Thanks for the updated app! One Q: is it possible to create an advanced user setting for “significant figures” to round tag values? I understand this may also be possible using math functions, but it would be much more convenient to have one global setting for it (especially for logging).
For example “significant figures= 3”: 1.234567 → 1.23, or 0.0056789 → 0.00568.
Yes. This app has a new app id. It’s not possible to get the same app id once the app got deleted from the store.
I don’t know. It’s working fine for me.
I can add a settings for that in the settings-page. But it will take some time.
Thank you, I sent you Diag. report if it helps - as all other apps works, so not sure where is the problem, I even restarted your app. I guess it’s not conflicting with old Insights trends I’m still using.
Maybe reinstall the app. This is a error on homey side not on mine. The diagnostic report looks fine.
OKi, reinstall didn’t help - let’s see if someone else will report the same. Thx.
It it possible that you use normal flows? Not adv. flows?
Ups, I didn’t notice it’s Adv. flow compatible only. Indeed, I tried to set/fix old Normal flows. So it really appear in Adv. flows but not in Normal flows.
Yes, because I used tokens for the cards. I would have to add two more cards for normal flows.
OK, because I saw it appear in the When, And … but not Then, so it was confusing. But if this is Adv. flow compatible only, everyone would have to rework their flows, which is a bummer. Anyway, maybe I’m the only case like this. Thank your for your replies.
I’ll add the other two cards soon. It’s not much work.
Added support for normal flows to the current test version.
@Sharkys now you have the then card on normal flows
@spkesDE I am having an issue with min/max. I simply want to calculate the energy consumption in the last 24h.
In Insights I see: Min=112, max =119
However, min is not correct:
The actual value 116.7 was measured around 12:40 today (about 6h ago).
Using 24 hours for the calculation gives the same result, and even 12h as well.
Any clue what could cause this?
there was a problem with my code. Forgot to add the timespan last 24h inside the code.
It has been fixed with the current test version.
That explains it then.
Works as expected now, thanks!