Data Curious 15.03.2019 - insight tools, more readable code, and the all-new DVS
data / curious
15.03.2019
Hello! It has Been A Minute.
Lots has happened in the data viz world over the past month or so. I do my best to pull out some highlights in this newsletter, but if you're looking for more frequent articles/resources find me on Twitter. I like to share things as I discover them throughout the week on Twitter, but only the best of the best make it into Data Curious.
This week I'm thinking about visualizations as insight-tools, how to explain complex theories visually, and the importance of writing readable code when analyzing data. I'm also very excited about the launch of the Data Visualization Society—if you haven't heard about it, we'd love to have you!
Read_
VR data viz, delivering insight, and the all-new DVS
How might VR change the way we see data visualization?
TBD. But Suzanne Borders of BadVR has some thoughts. Read this interview for a sneak peek.
Read on →
How can I deliver better insight through data visualization?
Read this interview with data viz pioneer Ben Shneiderman, in which he famously reiterates the mantra: "the purpose of visualization is insight, not pictures."
Work in data? Joined the Data Visualization Society?
You definitely should. Here's an article explaining why it was founded. Seriously, one of the most helpful, inspiring and interesting groups on the web.
Explore_
Data to sound, ternary plots, and explaining statistics
How can I turn my data into sound?
This tool lets you upload data, choose variables, instruments, speed and other features. Then listen to your data.
How can I visualize complex demographic data?
Take some inspiration from what I am now calling "the Brexit ternary plot". I personally love these triangular visualizations for their ability to capture so much data in one view.
All the angles →
How can I explain complex statistics and theories in a more visual way?
"Seeing Theory" is a visual, interactive introduction to probability and statistics.
Analyse_
Homelessness and wildfires
How has the US homeless population changed over time?
The HUD Homeless Point in Time Cound database contains "estimates of homelessness by state and estimates of chronic homelessness from 2007 - 2017". Variables include location, unique identifiers, counts and categories of homelessness.
Dive in →
What is the cost of wildfires in the US?
Stanford University has compiled a database from 2014 - 2017.
Learn_
Better Jupyter Notebooks, using Pandas with big data, and analysing Spotify data with clustering
How can I write more readable, reproducible code in Jupyter Notebooks?
You need this Jupyter extension in your life. Seriously. Read this, follow the steps, never write a terrible/disorganised notebook again.
How can I stop Pandas from freezing when I'm working with large datasets?
Try Dask: "...an Open Source project that gives you abstractions over NumPy Arrays, Pandas Dataframes and regular lists, allowing you to run operations on them in parallel, using multicore processing."
How can I analyse song data using clustering?
This Medium article is a nice introduction to clustering through Spotify data.
Spread the love.
Share this edition of Data Curious.
Say hello! I'm on: