of things people buy. Tags: django, twitter, dropletpay
After seeing a few tweets over the past few months pointing to people's Droplet purchases (ala the example below), I thought I'd map out the kinds of things people buy, and the amounts of money they spend.
— Craig Edmunds (@craigtech) June 19, 2013
For the uninitiated, Droplet is a way of paying for stuff using your phone. You charge up your phone with "credit" using your bank card, and you can pay for stuff simply and easily. I know a couple of the gents who founded the company who are based in my native Brum, so it's really cool to see the thing begin to spread.
So, to the hack. Last night I set about using the Twitter search API to find tweets mentioning @dropletpay, and matching a particular pattern ("I just paid [company name or Twitter handle] £x via @dropletpay"). I then parsed the tweet, pulled out the relevant details and stored them in a database. I then use the Twitter users API to get a bit more info on each company, like their full name, location, URL and description. The resulting map looks something like this (it's interactive, so you can click, pan and zoom):
(You'll need to read the article in full to see the map)
The data's really sparse at the moment as tweets drop off the search index after a week, so I've only got the last few days to play with, but it should update fairly rapidly so hopefully the leaderboard will start to become a bit more interesting.