Having issues getting Template to show "ok" "problem" i get "unknown"

[comet424]

ah ok so
availability: >
{{ ‘sensor.smoke_detector_2_by_wood_stove’ | has_value }}
means it checks for unknown or unavaliable if neither the avaliablity gets a value

availability: >
{{ states(‘sensor.smoke_detector_2_by_wood_stove’) | is_number }}
this checks for a number if it doesnt see a number its false but if it sees a number is true and avaliablty for a value is ok

is basiclly how i understand it…

are there docs that break down all the codes. that make it simple like for dummies as i assume this is all specific to home assistant and not to another coding language
only reason i say sometimes with my dislexia and learning disablity i find that the home assistant docs not always easy to understand sometimes not like a teacher teaching in class explaining things
for future knowledge

i appreciate the help as i only know basiclly templating and never know when you need to use avaliablity or when not to as i have templates that have avaliablity and some that dont …

wish this stuff was around in the 90s when i was programming PLCs in machine code this is alot easier now then back then doing lots of cool stuff

[petro]

@comet424

is basiclly how i understand it…

Yep

@comet424

are there docs that break down all the codes. that make it simple like for dummies as i assume this is all specific to home assistant and not to another coding language
only reason i say sometimes with my dislexia and learning disablity i find that the home assistant docs not always easy to understand sometimes not like a teacher teaching in class explaining things
for future knowledge

Yes

Templating

[SJ20035]

@petro

So, the “too long didn’t read” is:

using is_number on a states object will always work.

Thank you for trying to explain it and I do not really disagree with the use of is_number().

However, your logic is a bit flawed in that while those four integrations might use NumPy internally, Python has native float('nan') and float('inf') constants. You do not need NumPy to encounter these “objects” in a template.

It is true that this would be extremely rare but it could happen none the less - for example:

template:
  - sensor:
      - name: "Efficiency Calculation"
        unique_id: eff_calc_001
        state: >
          {% set solar = states('sensor.solar_production') | float(0) %}
          {% set load = states('sensor.house_load') | float(0) %}
          
          {% if load == 0 %}
            {{ float('nan') }}
          {% else %}
            {{ (solar / load) | round(2) }}
          {% endif %}

[comet424]

so does that mean you float 0 as nan or unavaliable as you dont use -1 and if it does get a value of 0 then pass it as a nan else calculate the value … is that how that works

[petro]

@SJ20035

However, your logic is a bit flawed in that while those four integrations might use NumPy internally, Python has native float('nan') and float('inf') constants. You do not need NumPy to encounter these “objects” in a template.

I did not mention those constructs because they still produce False with is_number. The only one that produces true with is_number is np.NaN. Try it yourself. Takes 2 seconds:

{{ float('nan') | is_number }}
{{ float('inf') | is_number }}

[SJ20035]

Thanks for making my point as you are correct that this is False:

@petro

{{ float('nan') | is_number }}

But my point is that the following would be True:

template:
  - sensor:
      - name: "nan_state"
        unique_id: nan_state
        state: >
            {{ float('nan') }}


{{ states("sensor.nan_state") | is_number }}

[SJ20035]

@comet424

so does that mean you float 0 as nan or unavaliable as you dont use -1 and if it does get a value of 0 then pass it as a nan else calculate the value … is that how that works

NaN stands for Not a Number.

In the world of computing and Home Assistant, it is a special value used to represent a result that is undefined or unrepresentable. It’s the computer’s way of saying, “I tried to do the math, but the answer doesn’t exist in the world of numbers.”

Examples of when this occurs:

• Divide by zero.
• Subtracting infinity from infinity.
• Square root of a negative number
• Trying to find the “Average” of an empty list (The Statistics integration does this often when sensors are offline).

[petro]

No it would not, because the state comes from the state machine.

Try it yourself. Very simple test.

EDIT: exact proof that you’re just making things up.

[petro]

@SJ20035

Divide by zero.

results in a ZeroDivisionError in python, will not occur anywhere.

@SJ20035

Subtracting infinity from infinity.

Possible, but also not possible because there’s no way to get inf in HA at this moment aside from you manually using it in Jinja to do this.

@SJ20035

Square root of a negative number

Will result in a ValueError unless using numpy.

@SJ20035

Trying to find the “Average” of an empty list (The Statistics integration does this often when sensors are offline).

No such function in HA. The jinja macro uses statistics.fmean which will result in a StatisticsError with an empty list.

Can we please move on now? All you’re doing is adding incorrect information to this thread.