playground/getting_tweets_senate.R

In gvegayon/twitterreport: Out-of-the-Box Analysis and Reporting Tools for Twitter

rm(list=ls())

source("R/text_mining.R")
source("R/verify.R")
load("data/senate_info.RData")

# Getting tweets from the senators
tmp <- sapply(twitter_accounts_senate, "[",1)
n <- length(tmp)
senate_tweets <- vector("list",n)
for (s in 1:n) {
  # Getting the info
  senate_tweets[[s]] <- tw_api_get_timeline(tmp[s],count=2000)
  if (!(s %% 10)) save.image("data/senate_tweets_example.RData")
  message(sprintf("%03d of %03d",s,n)," Senator ",senators$Name[s]," done...")
}

senate_tweets <- bind_rows(senate_tweets)
class(senate_tweets) <- 'data.frame'
save(senate_tweets,file="data/senate_tweets_example.RData")

# 
# load("data/senate_tweets_example.RData")
# 
# tweets_components <- tw_extract(senate_tweets$text)
# groups <- data.frame(
#   name      = senators_profile$tw_screen_name,
#   group     = factor(senators$party),
#   real_name = senators$Name,
#   stringsAsFactors = FALSE)
# groups$name <- tolower(groups$name)
# 
# senate_network <- tw_network(
#   tolower(senate_tweets$screen_name),
#   lapply(tweets_components$mention,unique),onlyFrom = TRUE,
#   group=groups, minInteract = 3)

# writeLines(tw_write_json_network(senate_network),'data/us_congress.json')
# 
# library(networkD3)
# forceNetwork(
#   Links = senate_network$links,
#   Nodes = senate_network$nodes, 
#   Source="source", Target="target",
#   Value="value",NodeID="real_name", Group="group",opacity = .9,
#   opacityNoHover = .5, legend = TRUE)
# 
# sankeyNetwork(
#   Links = senate_network$links,
#   Nodes = senate_network$nodes, 
#   Source="source", Target="target",
#   Value="value",NodeID="real_name") #, Group="group",opacity = .9,
#   #opacityNoHover = .5, legend = TRUE)