31:T835,**David, thanks for taking the time to share your story! I'm a big fan of your notebooks, especially the [Euler Spiral](https://observablehq.com/@dkirkby/euler-spiral?collection=@dkirkby/curves) one. Can you talk a bit about your background and your interests?**
I grew up in England and Canada, moved to the US for a PhD at Caltech, a postdoctoral fellowship at Stanford and I am now a professor at UCI. I started in the area of particle physics, working at [CERN](https://www.home.cern/) and [SLAC](https://www6.slac.stanford.edu/), then switched from studying the universe's smallest components to its largest scales, as a cosmologist. Today, I study the origin and evolution of the Universe with a particular focus on understanding the cosmic acceleration that drives the universe to expand faster each day. A lot of my day-to-day work involves data analysis, visualization, machine learning and astronomical instrumentation. **How did you get started in DataViz?**
I spent a summer in college working for an NGO in Kenya writing software to visualize field data in wood fuel use and conservation. **How long have you been working with data tools?**
My first data analysis project (in high school) involved punch cards so... several decades. **Which programming languages do you use and which do you like best?**
Over the years, my language of choice has evolved from C to Fortran to C++ to Python and, today, JavaScript is growing on me. **How much data do you typically work with?**
The astronomical surveys I work on produce 10s of terabytes of data but I usually work with a highly filtered subset. **How do you analyze, explore, and understand your data?**
Visualization, every single time. **What skills do you have now that you wish you had when you first started coding?**
Great question! Coding paradigms and ecosystems are constantly evolving so a lot of today's essential skills were not applicable earlier. So my answer is that I wish I had started with better intuition for reasoning about and solving complex computational problems.32:T1507,**Now, I'd like to learn a bit about how you use Observable. How did you first come across Observable?**
I have been a casual user of D3 for a long time but never really got over the hump to become proficient with it. I stumbled on some mentions of Observable while trying to solve a problem with D3 almost 2 years ago and was quickly hooked. **Why did you choose to use Observable for teaching?**
I discovered Observable right when I was faced with the challenge of teaching cosmology over zoom in Fall 2020. The potential for creating effective interactive explorations of abstract concepts was compelling and, as I discovered more of the ecosystem (coming from a mostly Python / Jupyter background) I realized it was also fun. **Where does Observable sit in your tool chain? How do you and your team(s) use it?**
After leaning heavily into Observable to create pedagogical material, I started to use it for my scientific research. It was not an obvious fit because my research community has invested heavily in Python development and Jupyter cloud infrastructure, and most of our data is in a file format (FITS) that is not well supported in javascript. However, I started with some problems where Observable shines and I am now gradually building out some the domain-specific scaffolding necessary for astronomy and cosmology applications. My first reaction after learning about Observable was "why didn't someone tell me about this earlier?" so I am now trying to [get the word out to my colleagues](https://observablehq.com/@dkirkby/desi-tutorial). I also use Observable for most of my personal projects now, including game AI engines and mathematical designs. **What are the top reasons you continue to use Observable?**
I still use Python and Jupyter for much of my research, where large datasets and compute resources colocated in the cloud are essential, but I believe Observable is the best choice today for many "edge" problems involving visualization and interactivity. But, more importantly, it is fun to use and the community is very engaged and supportive. **What would make your experience better?**
Most of the limitations I hit are intrinsic to javascript and the browser sandbox, and not due to Observable itself. However, what I miss most in the notebook is a way to drill down to any documentation for existing functions. JavaScript does not have an analog of the Python built-in help system, so perhaps there is an opportunity here for Observable to add significant value. In regard to collaborating with my data, most of the data I analyze is proprietary for some time before its public release. That limits collaboration via public notebooks. Observable Teams look like a good solution but it can be challenging to use research funding for online services. **Before using Observable, what data tasks weren't available to you, or weren't optimized?**
Two main things: adding powerful and intuitive interactivity, and sharing my work as a "live" document. **How does Observable help you in your data work?**
Observable enables the interactive visualizations that are essential to exploring and building intuition around complex multidimensional datasets and models. **How do you interact with the Observable community?**
I haven’t met other astronomers who are active in the Observable community but I have met many people solving similar problems in other domains. In that sense, I fit into the community horizontally rather than vertically, but this has a lot of advantages for cross-pollination of good ideas. **Has Observable lead to any interesting discoveries that may not have been possible otherwise?**
Not any ground-breaking scientific discoveries (yet) but I am making small discoveries every day with the tools Observable provides. **How does Observable’s repository of reusable code components help you?**
There is a growing library of tools that are sufficiently well-documented and general to fit right into my data workflow, and an even bigger body of notebooks that provide great inspiration or learning opportunities. **What other data tools do you use in conjunction with Observable? How do they work together?**
I rely on Python tools built on NumPy and SciPy for a lot of my numeric intensive data processing. I also use GPU-accelerated machine learning frameworks like PyTorch and JAX. I usually bridge these to Observable with CSV files or, for larger datasets, raw binary arrays. **What are some of your favorite notebooks that you've created, or that you're currently working on? Why?**
One of my favorites is [this one](https://observablehq.com/@dkirkby/penrose) to design infinite random quit designs based on Penrose tiles, because it was a challenging problem made easier with Observable. I had a similar experience with [this one](https://observablehq.com/@dkirkby/poly-collide), this time from my research on [fiber-positioning robots](https://observablehq.com/@dkirkby/desi-robot-motion-simulations), where a visualization revealed the inner workings of an algorithm. **Who inspires you in data visualization?**
I am very inspired by @rreusser's [WebGL visualizations](https://observablehq.com/collection/@rreusser/webgl) and @saneef's [algorithmic art](https://observablehq.com/collection/@saneef/re-plicas-mixes).33:[{"id":"e89f962186239313498ee404d49e8b7e2baf643cc07911c425b55cd03169a35e4c1e603c932f9cb32d2645f811b0c4e08be04ac09759360dea205d3449002919","href":"https://static.observableusercontent.com/files/e89f962186239313498ee404d49e8b7e2baf643cc07911c425b55cd03169a35e4c1e603c932f9cb32d2645f811b0c4e08be04ac09759360dea205d3449002919","name":"David.mp4","type":"video/mp4","lastModified":1657732293792.333,"size":3628893},{"id":"be8599d0c4b80d3491fc11eb687a49930feb172ac702a05aab5057a294d2b7d790ba6b579fa9350230960656f11c23eb06b89a641f20575be3458e08d017a39a","href":"https://static.observableusercontent.com/files/be8599d0c4b80d3491fc11eb687a49930feb172ac702a05aab5057a294d2b7d790ba6b579fa9350230960656f11c23eb06b89a641f20575be3458e08d017a39a","name":"David_Twitter@2.png","type":"image/png","lastModified":1657732293792.333,"size":117336},{"id":"40ef129cfbf6fcabf5de3b65a94e728500a07a71cb122a2d20da0157aa5d3ea1a474e0e47f306190102def3f5b425ab3ee5ad2413b82129c8bf6d6bfaef84729","href":"https://static.observableusercontent.com/files/40ef129cfbf6fcabf5de3b65a94e728500a07a71cb122a2d20da0157aa5d3ea1a474e0e47f306190102def3f5b425ab3ee5ad2413b82129c8bf6d6bfaef84729","name":"David-Notebook Cover_v2.png","type":"image/png","lastModified":1658434170475.838,"size":124218}] 2f:{"title":"Ambassador Spotlight: David Kirkby, Professor of Cosmology at UC Irvine","cells":[{"id":0,"value":"# Ambassador Spotlight: David Kirkby, Professor of Cosmology at UC Irvine","pinned":false,"hidden":false,"mode":"md"},{"id":143,"value":"

\n","pinned":false,"hidden":false,"mode":"html"},{"id":195,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":264,"value":"At the heart of Observable is the community. The people who make up our community create, collaborate, connect, and influence with data. The purpose of the Observable Ambassador program is to partner with community members who are actively using Observable. \n\nLearn more about the [Observable Ambassador program here](https://observablehq.com/@observablehq/ambassador-overview). \n\n---","pinned":false,"hidden":false,"mode":"md"},{"id":44,"value":"In June 2022, Professor David Kirkby was kind enough to share his story with the Observable team. We learned a bit more about his background and interests, as well as how he's used Observable in his research, in his classrooms, and...just for fun. \n\nFirst, some quick facts.\n\n**Ambassador since:** March, 2022
\n\n**Programming languages:** C, Fortran, C++, Python, JavaScript
\n\n**Uses Observable for:** Teaching, scientific research, fun
\n\n**Why Observable?**: _\"Challenging problems are made easier on Observable.\"_
","pinned":false,"hidden":false,"mode":"md"},{"id":250,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":237,"value":"$31","pinned":false,"hidden":false,"mode":"md"},{"id":197,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":203,"value":"

\n \n

David's infinite, random quilt design.

\n","pinned":false,"hidden":false,"mode":"html"},{"id":205,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":147,"value":"$32","pinned":false,"hidden":false,"mode":"md"},{"id":201,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":180,"value":"banner(\n \"Interested in learning more about collaborating with data?\",\n \"Sign up for free\",\n \"https://observablehq.com/product?utm_medium=social&utm_campaign=nba_data_jam&utm_source=observablehq.com\"\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":283,"value":"_____","pinned":false,"hidden":false,"mode":"md"},{"id":182,"value":"import { banner } from \"@observablehq/brand-assets\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":185,"value":"import { buttonCTA } from \"@observablehq/buttons\"","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W33","stdlib":"1","titling":"auto"}