3D Scatter with variable point size?

Is there a 3D scatter plot which allows the points to be displayed by size?

We did an experiment where we had people running at different speeds, up and down different inclines, and with different amounts of their bodyweight (60%, 80%, and 100%).
For each of these conditions, we measured how hard their muscles were working as a percent of their maximum.
Ideally I’d like to visualise this as a 3D scatter with the axes Speed, Gradient, Bodyweight, and then the points being the % activation of the muscles, but as different sizes.
I think this is a combination of Scatter:

And SandDance:

Any thoughts as to how to do this?
Perhaps could this be built in Charticulator?


If SandDance doesn’t do it for you, then I’m pretty sure this is not doable solely within Power BI, but if you’re willing to do some work in R and then bring those visuals into Power BI, you’ve got two excellent options in terms of R packages that can do this: plot3D and Scatterplot3d.

Here’s a video from BI Elite on how to do this in Power BI using the Scatterplot3d package. After that I’ve included some articles about Scatterplot3d and Plot3d, including the writeups by the respective package authors.

I hope this is helpful. I’m sure there’s also a way to do this using Python within Power BI, but I don’t know anything about Python…

  • Brian
1 Like

Thanks for the quick reply Brian, I really appreciate it.
I think there might be a Python solution using plotly as outlined here:

Something that was pretty close to what I needed was the Impact Bubble Chart

This may actually do the trick, but I need to dig deeper into the specifications and field wells as it’s not appearing entirely intuitively, nor does it seem to match up with the data.
Thanks again,


If that sort of bubble chart varying by color and size, with the play axis to provide the third dimension works for you, then you would be able to this natively (and relatively simply) within Power BI. Check out this video from Reed Havens that came out this week:

  • Brian

plotly Chart Studio was a pretty good solution here, and convinced me not to visualise the activation level using marker size: https://plotly.com/~rodw/1/

It’s great to know that you are making progress with your query @rodwhiteley. Please don’t forget if your question has been answered within the forum it is important to mark your thread as ‘solved’. Also, we’ve recently launched the Enterprise DNA Forum User Experience Survey, please feel free to answer it and give your insights on how we can further improve the Support forum. Thanks!