Alex Wein @alexwein@vis.social

Ooh just registered for Tapestry Conference. Anyone else going?

A small highlight of my day: a refactoring of our app's user model broke some downstream dashboards, leading to me telling everyone: "Do not use do_not_use!"

related to the last toot, Ira Glass's quote about the gap between one's taste and one's ability has been on my mind a lot over the last year, and it's been awesome to see it's resonated with other data folks:

Jenny Bryan had a great talk on code smells https://www.youtube.com/watch?v=91FQKciKfHI

@Allitorban introduces the IRA method: https://dataviztoday.com/shownotes/26

3 months into my new job, I've managed to convince a ton of people that I know what the hell I'm doing. Now I just gotta convince myself.

I'm in search of a good term to encompass relational data, regardless of technical implementation: rows of observations and columns of variables. The stuff in SQL databases and queries, data frames, sane excel worksheets and text files of all delimiters. I want something more technical than "rectangles" (I term I use a lot).

Grateful today and every day for ISO 8601

Clever and maybe useful barchart text function: https://gist.github.com/ltrgoddard/0f8d01bfc937fbcddef7204674b7f23d

sparkbar(c(1:5, 1:5, 6:10, 9:1))
[1] "▁▂▂▃▄▁▂▂▃▄▅▅▆▇█▇▆▅▅▄▃▂▂▁"

This article is pretty consistent with my experience as a Google TVC: https://www.bloomberg.com/news/articles/2018-07-25/inside-google-s-shadow-workforce

I've been listening *Make My Bed* by King Princess a ton over the last two weeks.

I'm still really new in my job, and gave a big presentation today on my vision for what I'm doing. I feel *very* clever for coming up with this kale analogy:

I admit a bit of aversion to non-interactive animation as a dataviz solution, but when the tools are as easy to use at gganimate, I'm sure I'll figure out some good uses.

Thomas Lin Pederson gave a keynote at UseR yesterday (big annual R conference, I'm not there) on the grammar of animation (via his new and improved gganimate package) and the slides are worth a look https://www.data-imaginist.com/slides/user2018/assets/player/keynotedhtmlplayer#1

And I gotta say, this was even easier to do with ggplot than anticipated. To go from the Line Chart on top to the Gutter Chart on the bottom was primarily one line of code.

It's not a perfect fit for data and chart type (it's very close to just a chart of y=x), and when and how to show nothing is a big and tricky problem. (And obviously the default axis labels aren't gonna cut it). Still, I'm excited about the idea (which is probably not completely novel).

One of my first vis.social hoots was a recommendation to read Understanding Comics. Here's a direct effort to use the grammar of comics in data viz, using gutters to omit the parts where nothing happens.