In dungates/TwitterStats: Twitter Sentiment Analysis

library(rtweet)
library(igraph)
library(hrbrthemes)
library(ggraph)
library(wordcloud2)
library(tidyverse)
library(twitteR)
library(qdapRegex)
library(tm)

Tweets over Time

Unfortunately the Twitter API limits scraping to just the last week, but there are a total of 743 engagements over the last week (including retweets, 319 engagements). Assuming a similar engagement rate over the previous 59 days we get about ~6000 total tweets including retweets, or ~2700 not including!

daysofdata <- search_tweets("#66DaysOfData", n = 5000, retryonratelimit = T, include_rts = T)
## search for 18000 tweets using the rstats hashtag
daysofdata %>% count(day = lubridate::floor_date(created_at, "day")) %>%
  ggplot(aes(day, n)) +
  geom_line() +
  labs(title = "#66DaysOfData tweets per day, 2020",
       caption = "source: Twitter Analytics, retrieved 2020-11-13", y = "Number of Tweets", x = "Week") +
  hrbrthemes::theme_ipsum_ps()

Wordcloud of Data

User Retweet Network

Basic network that labels the screen names of the most retweets over the #66daysofdata hashtag.

## create from-to data frame representing retweet/mention/reply connections

filter(daysofdata, retweet_count > 0) %>% 
  select(screen_name, mentions_screen_name) %>%
  unnest(mentions_screen_name) %>% 
  filter(!is.na(mentions_screen_name)) %>% 
  graph_from_data_frame() -> rt_g
V(rt_g)$node_label <- unname(ifelse(degree(rt_g)[V(rt_g)] > 20, names(V(rt_g)), "")) 
V(rt_g)$node_size <- unname(ifelse(degree(rt_g)[V(rt_g)] > 20, degree(rt_g), 0)) 

ggraph(rt_g, layout = 'linear', circular = TRUE) + 
  geom_edge_arc(edge_width=0.125, aes(alpha=..index..)) +
  geom_node_label(aes(label=node_label, size=node_size),
                  label.size=0, fill="#ffffff66", segment.colour="springgreen",
                  color="slateblue", repel=TRUE, family=font_rc, fontface="bold") +
  coord_fixed() +
  scale_size_area(trans="sqrt") +
  labs(title="Retweet Relationships", subtitle="Most retweeted screen names labeled. Darkers edges == more retweets. Node size == larger degree") +
  theme_graph(base_family=font_rc) +
  theme(legend.position="none")