(Part of the Guardian’s popular Prezi presentation of its data-journalism workflow)
… and well, it’s going to be awesome.
In case you missed my most recent email update, here’s the rundown of tomorrow afternoon:
WHERE: The Marketcrashers Hackernest, 231 Wallace Avenue, Toronto, ON
WHEN: Saturday, September 24, 1 pm - 4 pm
WHAT TO BRING: Laptop, power cord, and a determined, enthusiastic attitude
HOW TO RSVP: Please do so on this Meetup.
TENTATIVE EVENT SYLLABUS:
The workshop itself will have two “streams” happening simultaneously:
STREAM 1 (for hackers):
- A hackathon-style brainstorm-fest for the more advanced hackers
Like a hackathon, I’ll pick some specific datasets to hack on, as well as offer a prize for best project. (iTunes gift card? value ~ $100.)
Considering this is meetup is data-journalism-focused, the goal of the contest will be: Can you find and convey a story from this data? The finished project can be anything: an essay, a data-viz/infographic, an app, but it has to be web- publishable.
The non-competing “hacks” of the group act will act as as final jury on selecting the winner, since they have the narrative expertise and editorial sense to evaluate projects on their clarity, novelty, story cohesiveness.
Attendees would have the meetup time to brainstorm ideas and/or pick collaborators, and then have the following week to code the h*** out of it.
I will be sending out links to the candidate datasets and other rules of the contest to the RSVP list on Meetup, so be sure you’re on it if you plan to participate.
STREAM 2 (for hacking newbies):
A tutorial/skill-learning stream, with a planned step-by-step curriculum of exercises.
I was going to plan this around simple data-wrangling tricks I hear about through my job, but a data-hacker friend of mine made a really good point about journalists not trying to avoid coding if they really want to mine data.
In his words:
“To my mind, all of these various sites and tools are great so long as you have to have a problem / data set small enough to use with them … and have a problem that the tool is actually suited to address. For all of the effort spent, why not learn a fully general programming language and, having obtained mastery, wield great power over mere mortals?”
Frankly, I agree, and I think most journalists who want to hack are actually eager to learn, as long as they have some kind of curriculum to follow and someone to coordinate. So that’s what I’m going to help with.
So, the suggested general curriculum for newbies:
PART I:
- install Python before workshop
- intro to python syntax
- basic data types: primitives, tuples, lists, dictionaries
PART II:
- fetching a JSON file via HTTP
- working with JSON (much nicer than CSV if you can get it)
- storing data in a SQLite database
- querying the database
Tackling these face-on, from the ground-up, will empower you to:
- write and customize scripts - which is what data-scraping is all about.
- make full use of public APIs and actually know how to use the data once you get it
- be able to explore data more flexibly without relying on outdated, proprietary software like Microsoft Access
I’m getting Part I ready for this workshop, and maybe we can set up a followup workshop in a week or so to tackle Part II. Python is actually a really fun, easy language to learn, by the way. You’re going to love it!
And of course, everyone is free to discuss other projects they want to tackle, etc.
Again, if this sounds like something you’d like to join, please RSVP! (And if you can’t make it this time, please un-RSVP, so those on the wait-list can get in on it.)
-Momoko Price
Having said that, bringing businesses on board at the beta stage, let alone post-commercial launch, is no small feat for a funky, outside-the-box app like BuzzData. The concept of open data is still relatively new, and simple workflow tools for data wrangling and sharing are rare. Finding organizations that were hip to the movement and up for trying a new, untested digital app was a fun challenge, needless to say.
