Can you share this analysis across a longer time span?
For starters. There are so many possibilities here that any attempt to even think of a trend let alone what it's caused by, is wrong :) On the other hand: without knowing number of users, number of who's hring posts, when they were posted, location, ...: still really hard to give any sensible meaning to it. Just nice to look at.
I'm running the script from January 2018. However, covid was really disruptive, so I don't know what we'll be able to see. Anyway, I'll paste a link to the plot when it's done.
For starters. There are so many possibilities here that any attempt to even think of a trend let alone what it's caused by, is wrong :) On the other hand: without knowing number of users, number of who's hring posts, when they were posted, location, ...: still really hard to give any sensible meaning to it. Just nice to look at.