Good Graphs

Discussions about serious topics, for serious people
Post Reply
Allo V Psycho
Catbabel
Posts: 735
Joined: Sat Nov 16, 2019 8:18 am

Good Graphs

Post by Allo V Psycho » Sun Mar 28, 2021 4:26 pm

Split from the Bad Graphs thread as it is about Good Graphs.

OK, this is about Good Graphs, not Bad Ones. I thought there might be a thread on good data display, but can't find anything through the search engine. Mods feel free to move if I have missed a better thread. Done but please report stuff rather than post your request if you want us to read it - the only reason I ended up reading this post in the first place was because I'm doing an online shop and it's boring. Next time I might be doing a fun online shop and not end up reading all of the threads and your request could get missed. disco

I really like these two, but would value any guidance from the hive mind on how it is done

https://www.theguardian.com/world/ng-in ... -lockdowns

https://www.bmj.com/content/371/bmj.m3048

bagpuss
After Pie
Posts: 1695
Joined: Tue Nov 12, 2019 12:10 pm

Re: Good Graphs

Post by bagpuss » Mon Mar 29, 2021 9:14 am

I'm not sure how those particular graphs are done but you can do something somewhat similar to the Guardian one in Tableau (and probably the other one too), if it's the animation/timeline aspect you're interested in. The one thing I'm not sure about how to do is the way they've created the oval shaped mass of blobs where they should, presumably, all be showing a similar value and would be overlapping. I know how to stagger them to separate them out in a line up the otherwise unused x axis but they've got something a bit cleverer that allows them to make that oval shape. Mind you, I'm not entirely sure I like it as it makes it unclear whether a blob on the left is actually at a lower level than one on the right, or whether it's just pushed over there to make it fit. It's obvious from this chart, at the start when levels were mostly very low, that they're not all real values as there are a lot to the left of zero, for data where negative values are impossible, but it wouldn't always be obvious and so could be misleading.

Post Reply