How I came accross data…

Last year, I started working at my university’s bar, the Pic’Asso (“Asso” as association). I had the best position of all, responsible of IT.
Of course, like in most jobs on Earth, I had to do some repetitive tasks that I don’t personnaly like, that take time, and that are almost useless. Having to add articles to the register, or having to put videos on the tvs is not what I like doing the most. However, sometimes, and those are the rare times you love your job, I could do something fun, anything I wanted.

Some time ago, a friend of mine, and by chance the treasurer of the student’s halls (the Pic’Asso) at the same time, had the idea of simulating some sort of market law with our prices. He wanted all beers to cost more or less depending on their consumption; he called it The Wall Street of Beer.
So, obviously, we decided to start development. I began thinking of the algorithm needed to run the WS of Beer, and my colleague started to develop the calculator — in python :( . I wanted prices to rise as consumption rose, that way people that drank the same beers would be “forced” — by the prices — to test other beers, less bought, and thus less pricy.
Once we completed the algorithm and it was all coded, we had to show the new prices to the public, which was the hard part. At the time, we chose to make a table listing all our beers and their respective prices, just like in a stock market. We displayed the original price, the new price, the old price, the name, and obviously a green or red arrow along with the percentage of change.
This kind of representation was obviously easy to make, but very effective. People could directly see the price of their favorite and oh-not-so-favorite beers, and instantly choose which to buy. However one thing was missing,, they could not see a stock history of what had changed…
Some time later,
we hosted a big party in our university’s parking lot. We wanted to make it “transparent”, meaning that every student would know what they were paying for and what was being sold — the law in France makes it so any association must have a financial equilibrium of 0 at the end of every mandate. We only sold 3 kinds of beer, and all at the same price. However, we did not want them to know exactly how much money was being spent, but how much was being buyed. So we went for “one beer” as our base uni, and we decided to make a simple graph, which plotted the total number of beers sold every 10 minutes. And we added thresholds, to make it a little bit more interactive.
We displayed our graph in a big wall on the parking lot, which was right in front of the main stage where a DJ played his music. Everyone was able to see it, and everyone participated in its growth. It was beautiful.
6 months later
I started wornking on a project for a course at the uni. I had to find every startup born from the university, find its location, if it had any patents, how many workers it had, how much money it made…
We were only three working on the project, and our teacher wanted us to deliver a spreadsheet with all the information. Being computer engineers we obviously noted very early that we needed a database, and not a spreadsheet to better enter the information we found. Not only that, but our teacher had no idea how to use a DB, and even less SQL syntax. Also, our data had a lot, — and by that i mean a lot — of links. This company made transactions with this other one, and that one received money from that city, and this one periodically sold to the first one, and so on. We were in some kind of cul-de-sac.
So we began to think on a representation of the data, deviating from our real objective: getting it. We quickly thought of making graphs, they were perfect, representing links, and sizes, and colors. They could mean anything, but we had to make a thousand different graphs, for every different combination. However, our job is to think of “innovative” solutions, so we thought of making an interactive online graph, were the user could choose the data he wanted to see, and the colors to represent it. It was like magic, every color, every link and every bubble dancing on the screens.
What I do now
This was my introduction to data. I know search a lot of different representations of data online. I see how colors are used, what is pleasant and what is not, what can be seen and what can be heard, what must be interactive and what must not.
I also spend my free time searching for great APIs to get some more data, to mix it with other data, find links, show them, make nice graphs. I also love to find pre-existing representations, and try to make them easier to understand for the general public, or change this and that to make them reveal their hidden meanings.
I’m going to stop talking about my introduction to data, and how I’ve fallen in love with it, and let you see by yourselves:
- http://reddit.com/r/dataisbeautiful
- http://www.tylervigen.com/spurious-correlations
- http://informationisbeautiful.net
And I’m also going to give you the libraries I use to make my data: