Data Curious (2019.08.12): Yes, your chart still needs a good title
data / curious
2019.08.12
Greetings folks. Happy Monday (or maybe it's not, and that's ok too).
This week I'm thinking about finding inspiration in other's work, writing better chart titles, and visualizing machine learning among other things.
Is there something you've been wanting to read more about within data viz and data science? What's been on your mind lately? Drop me a note here and I'll see what I can dig up for some extra resources.
Sharing is caring, friends ✌️.
Read_
Who is your data viz hero?
Designer Tiziana Alocci started this lively Twitter thread last week of people listing the names of data viz practitioners who have inspired them most. Have a scan to learn about some names you haven't heard of before and be inspired.
Does my chart really need a title?
Yes. One thousand times, yes. This research summary article tells us why words still matter in data visualization.
How can we use new technology like virtual reality in data visualization?
Here's another question though: should we? Virtualitics is a company pushing heavily into this space. I found this article and teaser video interesting, but also a bit concerning: do we really want to release a bunch of 3d charts on executives at large? Feels like the technology is a bit gratuitous at this stage.
Explore_
Who serves up the best burger?
Nathan Yau has some ideas. Try not to be hungry staring at these burger-shaped radar plots comparing the taste profile of each major fast food chain in America.
How big is Android?
Big. This Bloomberg story has some mind-blowing graphics on how big the iPhone competitor has become. My favorite is the first one: a bubble plot shaped into a world map with proportional color shading and annotations?!! Maybe a bit overboard for some, but it's weird and I love it.
How much warmer is your city?
The BBC built a tool with spinning globes to visualize global temperatures and find out.
What does an exponential sum look like?
Prepare to get very nerdy with me for a second. Andrei Kashcha created a tool to visualize the output of exponential sums as a series of swirling, connected points. Stare into the beauty.
Analyse_
Where are all the icebergs in the world?
The Antarctic Iceberg Tracking Database contains data on icebergs from 1992 to present day.
Learn_
How can I create network maps for proximity analysis in Python?
Try this tutorial for using Pandana, a Python library for geospatial analysis. Nick Jones walks through his code to produce maps that analyze pedestrian accessibility in cities across the world.
How can I make plots to better interpret and explain machine learning models?
Check out these 4 Python libraries for visualizing machine learning outputs. Each has some reproducible code so you can get a feel for the workflow of each.
Spread the love.
Share this edition of Data Curious.
And say hello! I'm on:
Have a suggestion? Want to get in touch?
Fill out this quick form and I'll get back to you soon.
If you're not a subscriber yet to this newsletter, you can sign up .