date: 2014-02-27 22:12

I'm not really a Twitter pro, but after a few years with an account and periodic activity (I was and am still gunning for Diaspora ¹⁾, I had followed enough accounts to make my Twitter stream a little unwieldy, and this meant that I was missing out on tweets by people I'd actually like to pay attention to.

So, with a little time on my hands this evening (my family is out of town), I thought I'd play with the awesome twitteR ²⁾ package for the R Programming ³⁾ language. What I wanted to do, especially because the Twitter web interface is no good with this task, was list all those accounts that I follow but that do not follow me and all those that follow me and that I do not (yet) follow. Let's call this initial effort Tit for Tat Twit. And this really is just an initial run at the task. It can do for some optimizing — right now just working out the logic and getting more familiar with the twitteR package and Twitter's API.

library(twitteR)
 
# Set up Twitter authorization variables
# Requires creating a Twitter app: https://dev.twitter.com/
requestURL      <- "https://api.twitter.com/oauth/request_token"
accessURL       <- "https://api.twitter.com/oauth/access_token"
authURL         <- "https://api.twitter.com/oauth/authorize"
consumerKey     <- "keyHere"
consumerSecret  <- "secretHere"
twitCred        <- OAuthFactory$new(consumerKey = consumerKey,
                                    consumerSecret = consumerSecret,
                                    requestURL = requestURL,
                                    accessURL = accessURL,
                                    authURL = authURL)
 
# Initiate contact with Twitter
twitCred$handshake()
registerTwitterOAuth(twitCred)
 
# Get my account info; insert your handle in the quotes
tuser <- getUser("twitterHandleHere")
 
# Examine 'environment'
ls(tuser)
 
# Examine 'fields' in 'tuser' environment; just to test
tuser$url
tuser$name

Convert reference class 'lastStatus' into data frame with fields; only will return a data frame with one observation; just to test.

lsCsbStat <- as.data.frame(tuser$lastStatus)

Examples of methods available; look at environment to see more:

following <- tuser$getFriends()
followers <- tuser$getFollowers()
followers <- twListToDF(followers)
following <- twListToDF(following)

List my followers that I do not follow, and save as a data frame:

followers$Ifollow <- followers$screenName %in% following$screenName
followers$screenName[followers$Ifollow == "FALSE"]
noIfol <- as.data.frame(cbind(
    followers$screenName[followers$Ifollow == "FALSE"],
    followers$name[followers$Ifollow == "FALSE"]))
names(noIfol) <- c("Screen Name", "Name")

List following that do not follow me; save as a data frame.

following$noFollowMe <- following$screenName %in% followers$screenName
following$screenName[following$noFollowMe == "FALSE"]
noFollMe <- as.data.frame(cbind(
    following$screenName[following$noFollowMe == "FALSE"],
    following$name[following$noFollowMe == "FALSE"]))
names(noFollMe) <- c("Screen Name", "Name")

Same thing as first part above, just written differently.

I'm following them but they are not following me:

following$FollowMe <- following$screenName %in% followers$screenName
meNotThem <- subset(cbind(following$screenName,
                          following$name),
                    following$FollowMe == FALSE)

They're following me but I'm not following them:

followers$FollowMe <- followers$screenName %in% following$screenName
themNotMe <- subset(cbind(followers$screenName,
                          followers$name),
                    followers$FollowMe == FALSE)

The intersection; Not thoroughly tested yet:

meNotThem[,1] %in% themNotMe[,1] ||
    themNotMe[,1] %in% meNotThem[,1]

I used the resulting info to follow some folks that had followed me but for some reason I had missed them. I had followed a lot of news and other such types of accounts and I unfollowed those but kept them in lists so I can check on them every once in a while.

• categories:
  • Twitter
  • R Programming