I used the rtweet package to collect turfgrass community tweets from 2017. I had learned of this package from Bob Rudis’ blog post about rtweet.

At the end of the year, I had a bit of time, and a lot of curiosity, so I finally tried it. I wondered if I could find which turfgrass industry accounts had the most impact, or influence. To figure that out, I looked at these data:

  • the number of followers
  • how often the account tweeted (tweet creation rate)
  • how many times the account was mentioned by others
  • how many of the tweets sent from the account had a large number of retweets
  • how many of the tweets sent from the account had a large number of favorites
  • an overall ranking of the accounts, based on a combination of the ranks in those categories

The results of the calculations are in this searchable and sortable Turfgrass Twitter 2017 Shiny app.

The code I used is in the turf_twitter_2017 repository on GitHub, so you can see exactly what I did.

The 6,271 accounts I studied as “turf twitter” in this analysis were selected using these criteria.

First, I got the list of followers of seven turf scientists. I wanted to study accounts that are interested in new developments in turfgrass, and I figured that almost everyone interested in that topic would be following at least one of these accounts.

# get id of accounts following 
atc <- get_followers("asianturfgrass", n = 20000, retryonratelimit = TRUE)
pace <- get_followers("paceturf", n = 20000, retryonratelimit = TRUE)
unl <- get_followers("unlturf", n = 20000, retryonratelimit = TRUE)
tomy <- get_followers("striturf_tomy", n = 20000, retryonratelimit = TRUE)
jyri <- get_followers("Amplify_Turf", n = 20000, retryonratelimit = TRUE)
jk <- get_followers("iTweetTurf", n = 20000, retryonratelimit = TRUE)
unruh <- get_followers("jbunruh", n = 20000, retryonratelimit = TRUE)

Then I checked to see if the accounts identified in the previous search were also following one of the golf or sports turf industry associations.

# get accounts following industry organizations, which I'll use as a confirmation
# that the account is probably interested in turf
gcsaa <- get_followers("gcsaa", n = 20000, retryonratelimit = TRUE)
bigga <- get_followers("BIGGAltd", n = 20000, retryonratelimit = TRUE)
canada <- get_followers("GolfSupers", n = 20000, retryonratelimit = TRUE)
agcsa <- get_followers("AGCSA2", n = 20000, retryonratelimit = TRUE)
iog <- get_followers("the_iog", n = 20000, retryonratelimit = TRUE)
stma <- get_followers("FieldExperts", n = 20000, retryonratelimit = TRUE)

I identified the unique accounts in each list, and then did an intersect of the two lists to identify the accounts that followed at least one of the turf scientists and followed at least one of the industry associations.

# that's a lot of accounts, and many are duplicates
# to get the accounts to check, first select all those that are following one of the
# turf scientists

followers <- unique(rbind.data.frame(atc, pace, unl, tomy, jyri, jk, unruh))

# make list of unique accounts following the associations
assocations <- unique(rbind.data.frame(gcsaa, bigga, canada, agcsa, iog, stma))

# use the intersect function to select only those accounts that are in both follow lists
followers_assoc <- as.data.frame(base::intersect(followers$user_id, assocations$user_id))

colnames(followers_assoc) <- "user_id"
followers_assoc$user_id <- as.character(followers_assoc$user_id)

Then I removed the accounts that were set to private. I also cut those with less than 50 tweets. And I removed the accounts that were following 10,000 or more accounts themselves, as those appeared to be automated to increase followers.

# get basic data on these accounts
turf_follower <- lookup_users(followers_assoc$user_id)

# this removes the private accounts, also the least active ones
turf_follower2 <- subset(turf_follower, protected == FALSE & statuses_count >= 50)

# this attempts to remove those that seem to have automated follow/follower system
# in order to get a large audience, few removed here are turf related
turf_follower3 <- subset(turf_follower2, friends_count < 1e4)

That left 6,271 accounts for the further analysis.