Beautiful work Eli! I wonder when marks become a texture? Maybe even without being countable, the QDP dots afford frequency framing more than a solid PDF?
01.11.2025 22:05 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
I'll be at IEEE VIS in Vienna this week and I'm super excited for the great work our lab is presenting!
I'm on the job market as well -- come chat with me about affordances in visualization or tricky design problems you'd like me to (try to) fix :)
01.11.2025 22:03 โ ๐ 5 ๐ 0 ๐ฌ 0 ๐ 0
Racquel lays out the problem. You have a stack of density plots. Each curve has the same area (in pixels), but this means that they can have very different heights. This is a problem when you're running out of space.
You might be tempted to individually rescale each row so that each curve is equal height, but this would require changing the area (in pixels) of each curve, which can be misleading.
Instead of showing curves with matching heights, curves with matching areas support more accurate comparisons.
Example of equal area density plots from my unrelated paper ("Must Be A Tuesday"). If not for Racquel and Lace's insights, I probably would have made the "equal heights" mistake!
Stacked density plots should use the same y-scale between rows. Got a sneak-peak of this while plotting results for my paper, so I'm pleased to report Lace & @racquelf.bsky.social have already saved at least one designer (me) from hazards of irregular scaling! #ieeevis ๐
doi.org/10.31234/osf...
16.10.2024 17:34 โ ๐ 16 ๐ 5 ๐ฌ 2 ๐ 0
oh no I'm just seeing this now (still getting used to bluesky) -- thank you for the shout out Eli!! Glad our findings can be useful :))
21.11.2024 19:49 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
