Good morning, and welcome to The Graf. This newsletter is my attempt to, well, explain this newsletter.

It’s always difficult to translate thoughts into words, intent into action. This newsletter is my effort to do both; to write about a topic I love dearly (data visualization) on a more regular basis, and to actually translate my intention of being a great data journalist by, well, coding and actually building things each week. This newsletter is a vehicle, then, to hold myself accountable.

As I learn and grow and do, I’ll share everything with you; right here, in this newsletter, via this strange Internet-connected vehicle of thought sharing.

Thanks for reading.

A natural medium

It's not surprising that many new programmers are drawn to data visualization. The medium is visible, visceral; it is code with a tangible output. In many ways, I was drawn to data visualization for the same reasons. Algorithms seem abstract and hidden, in a way, whereas data journalism and visualization seemed an opportunity to blend storytelling with coding and reportage; a beautiful synthesis of creativity.

My first foray into code was in August 2018, when I took a bootcamp at the start of my PhD program at Caltech. By the following year, I had enrolled in a data science course. In March 2020, I heard about data journalism for the first time. Since that time — nearly two years now — my drive to be a data journalist has been unflinching. Love at first sight is bunkum, but this was love at first sight.

So I dropped out of the PhD and now I’m a data journalist at Spectrum, an autism news site in New York. Title of my job aside, there's still so much I need to learn. There will always be so much to learn. This may be because I’m constantly bombarded by beautiful work created by people more talented than I. And the number of skills that I’d like to learn — D3.js, more JavaScript, more data science, machine learning, natural language processing — feels unceasing and, often, insurmountable.

When confronted by such a huge amount of potential, and opportunity, it can be hard to take the individual steps, day by day, towards mastery in data journalism.

Consider this newsletter a promise to myself, and maybe to you, too.

I will improve, and learn something new about data journalism, each week.

Those words are a way to hold myself accountable, to say that I’ll take a little step each week and improve myself until I feel adequate as a data journalist; until I feel like I can do and build anything, even if it’s hard.

In cataloging and writing about my progress towards this goal, I hope that my journey illuminates new learning towards your own goals, too.

My goals are…

• To consistently read and follow data visualization research, grasping a deeper understanding of visual science and techniques to optimize someone's "intake" of a chart or graphic.

• To consistently study outstanding work by other data journalists; why does it work and how did they do it?

• To learn about people working in this field; to speak with them, and learn from them.

• To grow my own technical skills in each of the three (coding) pillars of data journalism: Collection, cleaning and exploring, and creation.

Technical trifecta

How does a data journalist turn numbers, documents or PDFs into an engaging, simple, approachable chart or graphic? 

The first step, I feel, is "collection." This includes, but is not limited to, the act of reaching out to retrieve a dataset, or create a dataset, perhaps using Python or R or JavaScript to scrape numbers from a website, or words from a cluster of PDFs. Collection, though, is deeper than this. It is the very act of story creation and ideation. This step is as much about knowing where to look for data, as it is knowing how to get it.

The second step is data cleaning and exploring. This is the wrangling and coercing and bending and shaping of data that data journalists do to *understand* a story; to get a feel for it. This is often accompanied with 'scratch' charts — simple graphics meant to peer more deeply into a dataset before any effort is made to create a compelling, final version.

The last step is creation. This is the synthesis of art and code — Python or R or JavaScript or D3.js or Plotly or Excel or Illustrator, or many others — to bend data into art. This is much more than just technical skills; it involves trained eyes, steady hands, careful consideration.

Each of these three pillars, even in isolation, is difficult to master. But they are worth pursuing, despite the monumental obstacles, because they help reveal the hidden. I believe that some of the greatest stories are out there, waiting to be told, if only a willing and capable data journalist would come along to understand the numbers, to synthesize them, to share their worth.

Expect this

• Regular roundups of notable, new research in data visualization. These are just the papers that I find, and study, and may or may not be all-encompassing.

• Regular roundups of exciting data journalism, from around the web. I aim to explain why I think they are exciting, or useful, or important, and will try to explain some technical tools used to construct them in the process.

• Regular updates on my own data visualization journey. In early newsletters, I might share something as simple as a scatter plot. But I hope that my charts — and the stories they tell — grow in complexity as this newsletter ages.

Thanks for joining me on this journey.

<3 Niko

This issue was written while listening to Ants From Up There by Black Country, New Road. Follow me on Twitter. Did someone forward you this email? Subscribe here.