- Joined
- Aug 1, 2013
This so perfectly illustrates Wu's Twitter addiction.I've been working a bit with R, a computer language used for statistical analysis of data. There is an add-on package that makes it easy to pull in Twitter data, and I wanted to share some quick insights about John's use of Twitter
Below is a histogram of the time spans between John's last 3200 public tweets, re-tweets, and replies (interactions). Unfortunately, the API limits me to the 3200 most recent interactions, and this data spans from Aug 21st, 2016 to October 23rd, 2016 -- roughly 2 months. The x axis represents minutes between tweets, and the y axis is the frequency across 50 bins or groupings. Upon initial examination, one can easily see it has a heavy right skew, meaning smaller time spans make up the majority of the data.
View attachment 146977
If you zoom-in a bit, it's actually becomes more clear
View attachment 146979
This data shows that of the most recent 3200 interactions John has had on Twitter, 485 (15%) have been within a 1 minute time-span of his previous interaction.
Some key points:
25% of John's interactions occur within 1.65 minutes of each other
50% of John's interactions occur within 5 minutes of each other
75% of John's interactions occur within 20 minutes of each other
154 (4%) have happened within 30 seconds.
Roughly 90% occur within an hour of each other and roughly 10% of John's interactions are more than an hour apart
The longest span John has gone between twitter interactions in this data-set has been 12 h
Only 45 (1%) interactions occurred more than 8 hours apart
3200 tweets and you can only capture two months of her Twitter use? Holy fuck.
Last edited: