How to combine multiple sensors into one new sensor for graphing?

Configuration Frontend
1

I have a sensor for each upcoming day, with the kWh forecast of solar production.
I’d like to show all days in a bar graph. Therefore all forecast sensors need to be combined into one, so apex charts (or others) can plot it.
Is there a way to combine multiple sensors into one? I would be fine with attributes, too, as there is a way with data_generator to extract them in apex charts.

My source sensors:

forecastSensors
forecastSensors1080×1144 93.9 KB

I tried to build something what I am looking for, by grouping each sensor at 24h rate. This gives me funky bars with lots of 0-bars in between:

grafik

The final result should look more like these blue bars:
grafik

2

Take a look at this topic, that came up 15 minutes before you posted. Was just coincidence, that I read your post first… :laughing:

3

Thanks for the link, but the topic wants to sum multiple sensors into one to get a total. Today that’s easy to do with sum helpers.

My question is on how to append multiple sensors into one in a timely manner. A form of concatenate, as each of the example sensors is for a different day.

4

I’m sorry, but then I don’t get what you want. :slight_smile:

You have the sensors shown in your screenshot above. How do you want these to be concatenated?

5

So you want it like your first graph but with thicker columns?

6

After realising, you’re speaking German, I kind of getting an idea what you want to do. :slight_smile:

I’d try it with a template sensor and set the days in the attributes. Is that what you need? Like in this example of a sensor:

template:
  - sensor:
      - name: solcast_something
        unit_of_measurement: "W"
        state: "{{ states('sensor.aktuelle_leistung') }}"
        attributes:
          prognose_day_1: "{{ states('sensor.prognose_day_1') }}" # heute
          prognose_day_2: "{{ states('sensor.prognose_day_2') }}" # morgen
          prognose_day_3: "{{ states('sensor.prognose_day_3') }}" # Tag 3
          [...] # und so weiter

This gives you one sensor, “solcast_something” with “Aktuelle Leistung” as state, and the upcoming days are listed as attributes from day_1 to day_x.

I hope this is at least the right direction. :slight_smile:

Let us know, if this fits your needs! :slight_smile:

7

yeai that’s it! :smile:

Maybe for better understandig of my initial post, I graphed my usecase with some ascii art:


sensor_forecast_day1: 11kWh  -----------+ 
                                        |
sensor_forecast_day2: 12kWh  -----------+---> sensor_combined_forecast
                                        |          - shows value of day 1 at timepoint now()+1day: 11kWh
sensor_forecast_day3: 13kWh  -----------+          - shows value of day 2 at timepoint now()+2day: 12kWh
                                                   - shows value of day 3 at timepoint now()+2day: 13kWh
                                            
                                            
graphed as timeline:
        11      12      13        kWh
 +-------+-------+-------+----...-->
now     day1    day2    day3

For my data, here is the complete solution I got so far:

template:
  - sensor:
      - name: "solar_forecast_timeline"
        state: "{{ states('sensor.solcast_pv_forecast_prognose_heute') }}"
        unit_of_measurement: "kWh"
        device_class: energy
        attributes:
            day_1: "{{ states('sensor.solcast_pv_forecast_prognose_morgen') }}"
            day_2: "{{ states('sensor.solcast_pv_forecast_prognose_tag_3') }}"
            day_3: "{{ states('sensor.solcast_pv_forecast_prognose_tag_4') }}"
            day_4: "{{ states('sensor.solcast_pv_forecast_prognose_tag_5') }}"
            day_5: "{{ states('sensor.solcast_pv_forecast_prognose_tag_6') }}"
            day_6: "{{ states('sensor.solcast_pv_forecast_prognose_tag_7') }}"

and the apexcharts series:

series:
  - entity: sensor.solar_forecast_timeline
    name: PV
    yaxis_id: kwh
    type: column
    color: '#FFBF00'
    show:
      datalabels: true
    data_generator: |
      return [
          [moment().add(1, 'days'), entity.attributes.day_1],
          [moment().add(2, 'days'), entity.attributes.day_2],
          [moment().add(3, 'days'), entity.attributes.day_3],
          [moment().add(4, 'days'), entity.attributes.day_4],
          [moment().add(5, 'days'), entity.attributes.day_5],
          [moment().add(6, 'days'), entity.attributes.day_6]
        ]

gives me this chart:

grafik

(blue: rain, gold: solcast, red and green: temps )

It screams to be treated as an array and loops, but I couldn’t figure it out :frowning:
Tried to adapt from

and

8

interesting, but is there a way to scale this up from 7 days to e.g. 2 years?

I’m trying to keep health records, e.g. weight, blood pressure etc

and sometimes, values like blood pressure requires several values taken together, systolic, diastolic and pulse