What do you do?

Until very recently, I was a postdoc and research associate in the School of Physics and Astronomy. I worked in Professor Lucy Fortson’s group, researching the evolution of galaxies and their central massive black holes. I was also the lead data scientist for the citizen science project Galaxy Zoo. I’m now a Fellow with Insight Data Science.

What tools/software/hardware/etc do you use to do your work?

Pretty much everything I do is on a computer. Most of my research involved large datasets (hundreds of thousands of galaxies, each with lots of associated parameters like mass, color, or brightness). The data we get from Galaxy Zoo is run by the Zooniverse, which runs on the massive resources of Amazon Web Services. Almost all my data analysis and visualization is done using Python, including lots of excellent tools specific for astrophysics in the astropy project. I like using Jupyter/iPython notebooks in our group as our version of the standard “lab notebook.” I edit everything in vi and write my papers in LaTeX; new sites like Authorea and Overleaf are making collaborative writing of proposals and papers much easier. Astronomy as a field is also lucky to have both an extremely high-use preprint server, the arXiv (meaning I worry much less about specific journal subscriptions), and an excellent interface for searching the literature called ADS. I read and manage my papers from arXiv/ADS in BibDesk.

Since my main research colleagues were scattered among several different timezones (Minnesota, Chicago, and Oxford), I’m a heavy user of Slack for both team meetings and informal chats. It’s replaced a lot of shorter conversations that used to exist only in email. All of our code and active development takes place on Github, and I’ve lately been putting more of my research materials there as well. When I publish a paper, I like having a single, open location on the web that I can point people to — it not only contains the manuscript, but all the raw data, figures, and code used to generate the results. Tools like that really help both open science and allow for reproducibility of your results.

What do you use that you love? What do you use that you wish worked better?

Slack is pretty great — it’s pretty simple, but vastly helps me feel like I’m part of a daily team even from thousands of miles away. I wish it was better at managing direct conversations (a continuous stream of posts in a channel can wander off topic), and I can lose track of whether something important was raised in an email, Github commit, or Slack thread. Github is also critical, but has a moderate learning curve and it’s sometimes hard to get colleagues to start using it.

I wish I had a better alternative to PowerPoint or Keynote. I have what are probably pretty high standards for what makes a good talk (in teaching, research or outreach), and I get frustrated a lot in talks where slide-based software isn’t used very well — mountains of text, illegible figures, distracting transitions, failure to practice, etc. I’m sure I fail at using it more often than I succeed, but I’m trying to keep improving. I’m always looking for better ways of effectively communicating what I do and how to do that at a variety of levels, including professional colleagues, students, and the general public.

Twitter: @kwwillett
Galaxy Zoo: www.galaxyzoo.org

---

Note from the librarian:

Congrats to Kyle on his Insight Data Science fellowship! For more workflow ideas, check out other posts in the How I Work series and visit the Productivity Tools page.

—Carolyn Bishoff