Questions/ideas #346
Replies: 5 comments 2 replies
I will try to fix this |
I will also try to fix this on the next update |
I'll solve it slightly differently than proposed. A value threshold would have two problems: it hides real data (you couldn't tell "no data / zero" apart from "small value"), and "small" is different for every sensor and unit, so everyone would have to tune it per entity. The actual problem isn't the value, it's the rendering: a small value on a large axis collapses into a ~1px sliver. So the fix will be a minimum visual bar height: any non-zero value is drawn at least 2px tall (matching the card's default line/point weights), while zero and missing data still draw nothing. You keep all three visual states, no bar (zero/no data), a 2px stub ("there is a value, but tiny at this scale"), and a normal bar and the tooltip always shows the exact number. No new option needed; it just works for every sensor at every scale. |

Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Hi @cataseven,
I have some ideas but would like to check if the possible features are already supported or can be realized by existing functionality before creating feature requests.
value_transform) but from my knowledge it is only possible to access data from the "own" entity and not from any other entity of the same chart. It would be great to also access data from other entities as this would allow to create graphs solely from data of other entities in the chart without a separate "own" entity (kind of virtual entity). For instance, you have 3 graphs in your chart and want to add a fourth graph which represents the sum of the other graphs (or some of the other graphs) without the need to create a helper entity for it. Then it should also be possible to omit a real entity for a graph.invert: truethe values are inverted on the chart (i.e., positive values are inverted and drawn in the negative area) but in the statistics legend, they are printed as is (i.e., not inverted). On the other side, MIN and MAX statistics are flipped, which is a bit strange. So, for instance if your sensor delivers values from, say 3 to 10, they are drawn in the area from -3 to -10 on the chart (correct) but in the legend, MIN is 10 and MAX is 3. It would be more understandable if MIN is -10 and MAX is -3 (applying the inversion also to the values) or MIN is 3 and MAX is 10 (completely ignoring the inverted flag here).All reactions