Superb, Erik!
I had to actually read the documentation (
) to figure out the meaning of “this period” / “last period”.
But it’s very neat!
The countdown-to-refresh indicator is brilliant!
Superb, Erik!
I had to actually read the documentation (
) to figure out the meaning of “this period” / “last period”.
But it’s very neat!
The countdown-to-refresh indicator is brilliant!
![]()
Both issues are probably caused by the same bug. The controls (containing the progress bar and legend) are rendered dynamically, and the chart should fill up the remaining space, but I think for you the chart is actually pushing the controls out of the viewport. I’ll try and reproduce and/or contact you privately to get this sorted.
I can’t make any promises yet, but I’ll see what I can do! Right now, I’m letting the chart library handle number formatting and precision while I focus on integrating it with Homey. I’ll need to check if I can make this work reliably, keeping potential edge cases in mind. I’ll get back to you on this.
I’ll see if I can add a setting to control this. I’m thinking of a padding option based on the data’s range. This way, you wouldn’t need to manually set min and max values and risk the chart overflowing when the data exceeds those limits. Does that sound like a solid solution to you? That said, the current scaling does give a pretty clear view of what’s happening! ![]()
Not by design so I’ll look into that as well. Thanks for reporting.
Thanks! I actually managed to reproduce this myself this morning. When you set the timeframe to “week,” the widget automatically displays Y-axis labels as days.
Since you selected the period “this” and today happens to be a Monday, there are only datapoints for Monday, which results in all auto-plotted Y labels showing just the name. This shouldn’t happen when all datapoints for the week are available—such as when selecting “last period” instead of “this period.” I’ll make sure to fix this so that it falls back to a more appropriate label in these cases.
Haha yes, those are spectacular
.
The padding solution sounds interesting. I am not sure how that will look like though and whether it would created ‘dead’/unused space above and below the graph. I do like the fact that it would auto-adjust to prevent overspills.
This may indeed cause unused space above and below. Another option might be to remove outliers or merge datapoints, but I’m not sure if I’m ready to dive down that rabbit hole just yet!
Another thought: would it be possible to let the user provide an optional min/max values set, which is disregarded by the widget if the series contains values that exceed the min or max?
Version 0.10.1 of DataVista is now available in the test channel! This version contains some minor improvements Line Chart Widget
:
More to come later, but I didn’t want to keep you all waiting ![]()
Good suggestion. I’ll take that into consideration as well, although this would probably result in the same unused space above and below, right?
Thank you again for this wonderful addition to your app!
I think it would be very interesting if the Y axis could be set to a variable value. Especially with PV information, you always have a idea of the values. Of course, there’s also a lot of dead space. But this can allso provide a certain level of information in various situations.
Good argument! I’ll see what I can do and how this should work when having multiple Y axes (I don’t want to have min/max settings per axis and overwhelm the widget settings).
Yes, but then the y-axis would cover the entire graph area and show that there are no datapoints in those areas. Whereas when using padding - at least in my understanding - the y-axis/graph is compressed and blank/padding space is added on top and at the bottom. On the other hand, I am making assumptions on how padding works
.
I realize now that using “padding” might not have been the best term to explain what I meant.
I was actually referring to adding padding to the min/max values of the y-axis based on the data range, not HTML padding around the chart. The idea is to extend the range slightly beyond the data points, which would add some space above and below the chart line (thus inside the chart itself).
However, I understand your concern, and I’ll definitely look into both options: adding padding to the range and manually setting the min/max values based on your and @Matthias_Konenberg arguments.
Thanks again for your feedback!
Ah, then I understand what you meant and think we are on the same page!
I also saw that the new version displays the unit now in the same color as the series color, instead of gray. Nice touch
.
The unit is indeed visualized in the respective color, but only when there are two series with different units or when using two Y-axes. This way, you can easily relate each series to the correct unit/axis.
For a single Y-axis, the unit remains gray to avoid unnecessary distraction and keep the focus on the data itself.
Ah, when having 2 series and one Y axis it is gray, but colored when having a single series. That’s a bug and I should align that. Thanks!
Thanks Erik, for this great app.
Wonderfull !
I just started to use it and was playing with the simple gauge to show the P1 total power value. In my case this can be negative (solar panels) as well.
First I entered values for minimum and maximum levels, but now wanted to clear the range to let the app calculate it. But that is not possible.
When I clear the earlier entered value it says “Invalid value ()” when I want to return. So I am stuck…
I use the latest test version of your DataVista app (0.10.1).
Hi Bernd! Thanks for reporting, I’ll try and reproduce and get a fix in place. For now, I guess the easiest way forward would be to remove the widget and add a new simple gauge, leaving the min/max values empty from the start.
Hi @Erikvl87
Update:
Both graphs shows the inside temp. The upper incorrect. The lower correct. Same ‘databron’. It looks like it just changed the scale.
Yesterday the scale of the inside and outside temp were the same
Rogier
Hi Rogier, I think I’m missing the issue on your first screenshot. The inside temp (orange) from your first screenshot is tied to the Y axis on the right and looks aligned.
If the ranges of both datasources are to far apart from eachother, the widget decides to render a second Y axis.
To prevent this from happening, you may want to update to version 0.10.1 and use the new widget setting Y Axis Calculation Method and set it to Force same axis.