Hard to tell if the data is wrong if the time axis is not the same. Seems to be only a very small portion. In that case the automatically chosen y axis range is different, causing small variations to appear much larger.
Same issue here. After upgrading to the latest HA version I’m getting strange results. Max value is totally wrong. I also believe it has to do with the DB changes.
No the data is completely wrong. These graphs were working fine until the latest updates
How these graphs are supposed to be compared?
The stock history-graph shows (supposed to) real values.
The mini-graph-card
shows approximated values - and this approximation depends on parameters like smoothing
, points_per_hour
, aggregate_func
.
You even have not provided showing Y-labels on the 1st graph and same hours_to_show
.
To perform an analysis, you should ensure same settings for both cards.
I’ve been running these graphs for months. I know what they are supposed to look like. The daily energy graph should look like a staircase going upwards and it resets to 0 everyday. I get that you can’t compare these exactly but the point was that something has changed and it’s obvious from the min and max values.
This is showing the same data in the stock graph and mini-graph-card. Obviously the mini-graph-card is wrong. Ther mini-graph-card shows a steep increase. I believe this happened when I upgraded to 2022.6.1
cards:
- entities:
- entity: sensor.lumi_6d760604_temperature
extrema: true
font_size: 58
height: 350
hour24: true
hours_to_show: 24
index: 0
line_color: blue
line_width: 4
name: Eingang
points_per_hour: 60
type: custom:mini-graph-card
How to compare mini-graph-card
& history-graph
:
Some people are experiencing a bug now.
To compare graphs - use this card.
(it uses the same method as was used here)
- Create 4 helpers:
input_number:
test_compare_graphs_padding_left: &ref_padding
min: 0
max: 200
step: 1
mode: box
test_compare_graphs_padding_right: *ref_padding
test_compare_graphs_padding_top: *ref_padding
test_compare_graphs_padding_bottom: *ref_padding
-
Create a “panel” view.
-
Place this card on a view:
type: vertical-stack
cards:
- type: entities
entities:
- entity: input_number.test_compare_graphs_padding_left
name: padding-left
- entity: input_number.test_compare_graphs_padding_right
name: padding-right
- entity: input_number.test_compare_graphs_padding_top
name: padding-top
- entity: input_number.test_compare_graphs_padding_bottom
name: padding-bottom
- type: custom:mod-card
card_mod:
style:
hui-vertical-stack-card $: |
div#root {
display: grid;
}
div#root > * {
grid-column-start: 1;
grid-row-start: 1;
margin: 0px;
--ha-card-background: rgba(0,0,0,0);
}
card:
type: vertical-stack
cards:
- type: custom:mini-graph-card
entities:
- entity: sensor.xiaomi_cg_1_pm25
show:
labels: false
name: false
state: false
points: true
line_width: 1
hours_to_show: 1
points_per_hour: 120
update_interval: 0
aggregate_func: last
smoothing: false
card_mod:
style: |
ha-card {
padding-left: {{states("input_number.test_compare_graphs_padding_left")}}px !important;
padding-right: {{states("input_number.test_compare_graphs_padding_right")}}px !important;
padding-top: {{states("input_number.test_compare_graphs_padding_top")}}px !important;
padding-bottom: {{states("input_number.test_compare_graphs_padding_bottom")}}px !important;
}
- type: history-graph
entities:
- entity: sensor.xiaomi_cg_1_pm25
hours_to_show: 1
refresh_interval: 0
-
Specify your own monitored sensor instead of “
sensor.xiaomi_cg_1_pm25
” and a required frequency: since my sensor is updated every 30 sec, I am using “points_per_hour: 120
”. -
Use input helpers to align graphs together.
Note that the mini-graph-card
is updating and “scrolling” a graph to the left.
IMHO the history-graph
should also automatically update and “scroll” to the left - but it does not seem to do it now (due to a recent bug I think).
After every update the alignment is lost, so you should re-align, make a screenshot and then analyse it in MS Paint (or whatever).
It’s not about comparing graphs. The data in the Mini graph card is simply wrong. Look at my graph above. 93.5 degrees celcius max!
This used to work correctly for a long time. Only after the HA upgrade it’s wrong.
- The proposed solution is not intended for YOUR personal case only. It is for comparing cards in general; for analysing a difference between different “
aggregate_func
” values etc. - As for YOUR personal case - check the GitHub issue, some people sorted out a reason of their similar problems.
P.S. … not to mention errors in your code…
One more picture w/o a code, come on))
Start with this:
- restart HA (or the whole HA OS).
- restart DB if it is used in a separate container.
- clear cache.
Here is
Points 1 and 3 implemented
I don’t know how to restart the DB, I use HaOS on Rpi built-in DB
The only thing that preceded it was the exclusion of some entities from the recorder that had nothing to do with temperature.
type: vertical-stack
cards:
- type: custom:mini-graph-card
entities:
- entity: sensor.venkovni_teplota
color: green
show_state: true
show_points: true
hours_to_show: 1
points_per_hour: 120
aggregate_func: last
update_interval: 0
height: 80
show:
state: true
points: true
extrema: true
name: false
icon: false
line_width: 1
hour24: true
lower_bound: ~0
smoothing: false
- type: history-graph
entities:
- entity: sensor.venkovni_teplota
hours_to_show: 1
refresh_interval: 0
I rebooted the host. It seems it’s now working fine again. Restarting HA did not help.
Could you specify a type of your installation?
I am using Debian + HA in Docker (with a native DB). Rebooting the HA causes rebooting DB server too.
I’ll restart the host one more time and let me know
Thanks for your help. I use raspberry pi with latest OS (bullseye) and HA Docker with default DB.
IMHO the situation is strange.
Yes, this HA version has some issues with Recorder, history-graph & Logbook.
From other side, it is strange that - w/o any reboots - history-graph
shows SOMETHING and mini-graph-card
does NOT.
Whose fault it is? Bugs in HA? Or mini-graph-card
is just NOT ADAPTED to some new requirements, I don’t know…
Then it is strange that rebooting HA (all-in-one) could not help.