Python beginnings — ChiPy Blog #1
My forays into coding basically run as long as my education, but let’s start with a more recent rendezvous. So before I get into this first blog post for the ChiPy mentorship program, let’s have a throwback Thursday moment.
About 2 years ago, I fell deep in love with all things data journalism (design, coding, building web apps, the works). I was fortunate enough to be a part of the first cohort of the ProPublica Data Institute in the Summer of 2016. Before that, I wasn’t even aware that folks were making careers out of digging through large datasets and wrangling them into simply beautiful and interactive designed websites.
I’d found my particular brand of nerd hive. So I had to dig deeper. I made a list of things to try.
I got a job working on interactive health programs on the web, but I wanted to dig deeper. I wrote a little JavaScript on the job and stretched my HTML skills. Then I signed up for an R for Journalists class and got a certificate. And I applied for a ChiPy Python mentorship and actually got it!
I’d dabbled with Python before in an attempt to automate some processes at work, but I still considered myself somewhat of a novice. So I struggled a little with picking a project. I felt a bit pulled in two directions — on one hand, I wanted to learn how to automate process and flex a coding muscle. On the other hand, I wanted to build shiny, pretty pinnacles to data.
Luckily, my focus came in a convenient and clear form, and again from ProPublica.
Picking a project and getting started
After the mentorship has its kick-off dinner, I went to ChiHackNight to hear a presentation by ProPublica Illinois about how their treasure trove of data on Chicago parking and vehicle ticketing. They’d already published a few big stories on city sticker tickets with the data, but were in need of people who wanted to dig deeper and find other insights. Thus, a Python project was born.
That night, I worked with a group of people to fork a sample of the data from GitHub and get Python 3 installed properly on my laptop. I previously had a foray with the Anaconda distribution that had a lot of conflicts with a fresh install of Python, but I had a ton of help getting Anaconda off and a fresh install of Python on. The next week, I met with my mentor, Ed, to get Python 3 fully up and running with the rest of what I would need for my project — the virtual environment, Jupyter notebook, and Atom. It was really helpful to have someone to guide me through the somewhat frustrating process of getting all of this set up correctly. I imagine if I had been struggling with it solo I might have thrown my computer out of the window.
Once I got my own Jupyter notebook set up, the following week was dedicated to exploring the sample data. Since I had some experience with the data wrangling concepts from my R class, it was extremely helpful when Ed pointed me toward the Pandas cheat sheet! Here was a list of all the data manipulation concepts I’d learned, but in a new language. I finally had the tools at the ready to really dig into the data and find something interesting.
So what’s my plan moving forward?
This week, I’m finishing up my exploration of the Chicago sample tickets data. Once I find some notable answers to my questions, I’ll load the full dataset and run some of the same analyses on it to get a fuller picture of what I’m seeing.
I’m starting with these resources:
After analysis
Once I move beyond data exploration, I want to build a well-designed web app for a user to explore the data and interact with it in a way that tells a story. My goal is to have a beautiful and engaging site to present my findings.
My next post is due in October, so I hope to have wrapped up analysis and moved on to building the web app by then.
I’m VERY excited for this journey!
