I am new to the rtweet package and have versions of the same question related to queries passed to POST statuses/filter.
Per Twitter, "The default access level allows up to 400 track keywords, 5,000 follow userids and 25 0.1-360 degree location boxes." Does stream_tweets follow this structure? For example, could I pass 200 keywords and 300 user IDs and 2 bounding boxes? If yes, how does stream_tweets know which parts of the passed vector are keywords, user IDs, and bounding boxes?
The documentation of stream_tweets, for example, says that a bounding box "must be provided via a vector of length 4, e.g., c(-125, 26, -65, 49)". I interpret this meaning only one bounding box per connection, which is less functionality than POST statuses/filter allows. Same if stream_tweets does not distinguish between user IDs and keywords. I believe I would just pass everything as part of one long character vector, e.g. 'car,table,box,3892190,38901930,-74,40,-73,41' would give tweets with any of those 3 words, from any of the 2 users, or from within that last bounding box.
