Forum Discussion
Smooth lines
hi. there is not a way to do this in powerbi that i am aware of.
I know that the smoothed line looks good but is it an accurate representation of the data?
hi. there is not a way to do this in powerbi that i am aware of.
I know that the smoothed line looks good but is it an accurate representation of the data?
"..is it an accurate representation of the data?" - it can well be more accurate than showing only the measured points..
In case the data in question is anyway an estimate and contains error, representing absolute points instead of smooth lines indicating that "it is roughly like X" is totally and utterly warranted in my opinion. Also, for cases where you have continous phenomenon underneath (such as temperature...) but have only discontinuous measurements (say hourly), again, smoothing the lines to be continuous more truly represents the underlying distribution, of which the collected data is only a partly representing sample. I disagree with your stance for many a dataset of reality.
If you want to represent a "Gauss Bell", it is completely necessary to have smooth lines to understand the distribution.
That's because i think it is necessary to have the possibility to represent a smooth line chart.
Josep.
This was a while back...any chance this is possible in new version?
Seems like a terrible idea anyway. In that example picture the curves make the lines go higher and lower than the actual data points they're connecting. So the graph is lying. It looks pretty but it's worse than useless for representing data. It's a good thing that we don't have this "feature".
If you do want this feature here you go
https://zoomcharts.com/en/microsoft-power-bi-custom-visuals/
"..is it an accurate representation of the data?" - it can often well be more accurate than showing only the measured points..
In case the data in question is anyway an estimate and contains error, representing smooth lines instead of absolute points only indicating that "it is roughly like X" is totally and utterly warranted in my opinion, being more truthful representation of the data. Also, for cases where you have continous phenomenon underneath (such as temperature...) but have only discontinuous measurements (say hourly), again, smoothing the lines to be continuous more truly represents the underlying distribution, of which the collected data is only a partly representing sample. I disagree with your stance for many a dataset of reality.
