edge_tweets.R

# load required packages
library(rtweet)
library(tidyverse)

# get tweets (you will need a Twitter API token to run this code)
edge <- get_timeline("edge", n = 3200)

# function for frequency table
count_pct <- function(df) {
  return(
    df %>%
      tally() %>% 
      mutate(n_pct = 100*n/sum(n))
  )
}

# gender of tweets about Edge final question
questions <- edge %>%
  filter(grepl("asks his final|asks her final|asks their final|ask their final", text, ignore.case = TRUE)) %>%
  mutate(gender = case_when(grepl("asks his", text) == TRUE ~ "man",
                            grepl("asks her", text) == TRUE ~ "woman",
                            grepl("asks their", text) == TRUE ~ "neutral singular",
                            grepl("ask their", text) == TRUE ~ "neutral plural"))

questions_gender <- questions %>%
  group_by(gender) %>%
  count_pct()


# top mentioned accounts
mentions <- edge %>%
  filter(!is.na(mentions_screen_name)) %>%
  select(mentions_screen_name) %>%
  unnest() %>%
  filter(mentions_screen_name != "edge") %>%
  group_by(mentions_screen_name) %>%
  count_pct() %>%
  arrange(-n)

# top retweeted accounts
retweets <- edge %>%
  filter(!is.na(retweet_screen_name)) %>%
  select(retweet_screen_name) %>%
  filter(retweet_screen_name != "edge") %>%
  group_by(retweet_screen_name) %>%
  count_pct() %>%
  arrange(-n)