No one has ever asked me for help in Python before
Seriously though, the only Python I know is from using HA although I did also start with Fortran 30 something years ago.
My Up time sensor looks like this which was made better by @khan3962’s hint about using ‘%’
rpi_uptime:
friendly_name: RPi Uptime
entity_id: sensor.time
value_template: >
{% set up_time = as_timestamp(now()) - as_timestamp(states('sensor.last_boot')) %}
{% set minutes = (up_time // 60) | int %}
{% set hours = (minutes // 60) %}
{% set days = (hours // 24) %}
{% set weeks = (days // 7) %}
{% set minutes = (minutes % 60) %}
{% set hours = (hours % 24) %}
{% set days = (days % 7) %}
{% macro phrase(value, name) %}
{%- set value = value %}
{%- set end = 's' if value > 1 else '' %}
{{- '{} {}{}'.format(value, name, end) if value | int > 0 else '' }}
{%- endmacro %}
{% set text = [ phrase(weeks, 'week'), phrase(days, 'day'), phrase(hours, 'hr'), phrase(minutes, 'min') ] | select('!=','') | list | join(', ') %}
{% set last_comma = text.rfind(',') %}
{% if last_comma != -1 %}
{% set text = text[:last_comma] + ' and' + text[last_comma + 1:] %}
{% endif %}
{{ text }}
Feel free to ask any questions bearing in mind what I said about not being a Python expert
Thank you so much for your quick response and your help! I’ve put your code in my sensors.yaml and it works so much better…well, 20 minutes since my host reboot, anyway
You may not consider yourself a Python expert but you’re way ahead of me! And I also stole your mdi icon, it’s nicer than the boring clock I had before
I’ve stolen these templates from you! Hope you don’t mind
Any idea why I might be seeing the following results? I’m not familiar with a couple of things used here so I can’t really figure out what’s going on. (specifically macros in jinja, phrase(), and ‘{} {} {}’). Looks like I’ve got some googlin’ to do but was curious if your values are correct because this was a direct copy/paste.
As far as I can tell it’s interpreting the value here as days instead of hours maybe?
In this screenshot sensor.uptime is showing an accurate value. sensor.rpi_uptime (the stolen template) is showing a very incorrect value.
@DavidFW1960 Like I said originally, I so believe the ha_uptime is interpreting sensor.uptime as days. Anyone can see that mathematically. I didn’t know why.
sensor.uptime was the correct value ha_uptime was incorrect I’m pretty sure I know how long ago I rebooted my system considering it was only 15 minutes ago.
So to answer my own question. I have the uptime sensor set to report in hours. I would bet my bottom dollar @klogg has his set up for days (the default) and that if I change mine to match all will be good with the world.
- platform: uptime
unit_of_measurement: hours
Thank you for your input though.
Edit: That sounded snarkier than I meant. I do appreciate you looking at it.
I’m not sure what version you are using here (there have been a few iterations) as the quoted snippet doesn’t format properly, but I currently use this:
system_uptime:
friendly_name: System Uptime
entity_id: sensor.time
value_template: >
{% set up_time = as_timestamp(now()) - as_timestamp(states('sensor.last_boot')) %}
{% set minutes = (up_time // 60) | int %}
{% set hours = (minutes // 60) %}
{% set days = (hours // 24) %}
{% set weeks = (days // 7) %}
{% set minutes = (minutes % 60) %}
{% set hours = (hours % 24) %}
{% set days = (days % 7) %}
{% macro phrase(value, name) %}
{%- set value = value %}
{%- set end = 's' if value > 1 else '' %}
{{- '{} {}{}'.format(value, name, end) if value | int > 0 else '' }}
{%- endmacro %}
{% set text = [ phrase(weeks, 'week'), phrase(days, 'day'), phrase(hours, 'hr'), phrase(minutes, 'min') ] | select('!=','') | list | join(', ') %}
{% set last_comma = text.rfind(',') %}
{% if last_comma != -1 %}
{% set text = text[:last_comma] + ' and' + text[last_comma + 1:] %}
{% endif %}
{{ text }}
And in case you’re interested (although I’ve seen some of your work on here so I doubt you need it) this is the Lovelace:
Hey, that looks like the template I helped @Mariusthvdb with . Anyways, I’ve since moved away from that towards a simpler solution. But it comes with drawbacks. I use the timestamp device_class which makes the frontend do all the week/day calculations. The drawback is that it only shows the ‘main’ largest value. I.e. if its 8 weeks and 4 days, you’ll only get 8 weeks ago.
pros:
No need for sensor.time as an entity.
Less load on your system because of point 1.
Cons:
Only get ‘highest values’.
Some custom cards don’t ‘translate’ the timestamp and it requires heavy JS.
I actually prefer the highest value thanks for sharing. I have sooo many templates in my config. I could keep you busy for a month simplifying them I reckon I could reduce my metric fuck ton of sensors by a third with your help lol.
I’m the same way, I have a ton that can be simplified but I just haven’t until there’s a good reason to. I have some templates that are 4 years old at this point. If it works, don’t fix it!
This is one example that I did update though because of system load.
