smile_emoji_twitter

library(rtweet)
library(dplyr)
library(tidytext)
# the q setting is a single emoji, but I got a unicode spacing issue when writing it:
happy_face <- search_tweets(q=" 😀", n=100000, token=bearer_token(), retryonratelimit = TRUE)

happy_centric <- happy_face %>% 
  filter(is.na(retweet_status_id) & is.na(quoted_status_id)) %>%
  select(status_id, text) %>% unnest_tokens(char, text, token="characters") %>%
  group_by(status_id) %>% mutate(before=lag(char), after=lead(char)) %>% ungroup() %>%
  select(before, char, after) %>% 
  filter(char == "😀")

either_side <- data.frame(char = c(happy_centric$before, happy_centric$after),
                          place = c(rep("before", length(happy_centric$before)),
                                    rep("after", length(happy_centric$after))),
                                    stringsAsFactors=FALSE) %>% filter(!is.na(char))

either_side %>% count(place,char) %>% group_by(char) %>%
  summarise(prop_before = sum(n * (place == "before")) / sum(n), number=sum(n)) %>% 
  filter(number > 30) %>% View()