After posting my 10.000th tweet on Twitter, I downloaded all of them and did some data analysis in Python. This blog post includes 10 data visualizations from analyzing 10.000 tweets.
Note: visualizations were created using Python libraries Matplotlib and Seaborn. The Jupyter notebook can be found here.
1. Tweets by Type
Starting with a very basic visualization showing the total count of tweets per type.
2. Tweets by Source
During my almost 10 years on Twitter I’ve been using a lot of different sources (i.e. Twitter clients).
3. Tweets by Type per Year
This time series visualization makes it easy to determine my most (in)active years on Twitter.
4. Mean Tweet Length per Year
Correlation and Causality aside - can you guess in which year Twitter increased the maximum number of characters per tweet?
5. Tweets per Year/Month
Another way of visualizing time series data.
6. Tweets by Source per year
This is my favorite: so many different plots but still easily readable! Also: I miss twicca!
Wordcloud consists mostly of Croatian words since I used to tweet almost exclusively in Croatian up until a few years ago.
8. Tweets per Weekday
No major differences between weekdays - I like to tweet every day!
9. Tweets per Month Day
Similar as before, but for month days.
9. Tweets per Hour
Last but not least, a visualization showing number of tweets sent per hour of the day.
And that’s it! If it hasn’t been clear by now, I’m very active on Twitter (user: tgel0) and I would love to hear from you! Send me a tweet and let me know what do you think about this blog post.