title: "Twitter Coverage of the useR Conference 2018"

author: "Neil Saunders"

date: "`r Sys.time()`"

github_document:

html_document:

keep_md: yes

number_sections: yes

toc: yes

toc_float: yes

```{r setup, include=FALSE}

knitr::opts_chunk$set(echo = FALSE, message = FALSE, warning = FALSE)

library(rtweet)

library(tidyverse)

library(lubridate)

library(knitr)

library(pander)

library(wordcloud)

library(tidytext)

library(igraph)

theme_set(theme_light())

user2018 <- readRDS("../../data/user2018.rds")

```

An analysis of tweets from the useR 2018 meeting. `r nrow(user2018)` tweets were collected using the `rtweet` R package:

```{r search-twitter, eval=FALSE, echo=TRUE}

library(rtweet)

user2018 <- search_tweets("#useR2018", 10000)

```

```{r tweets-by-day}

user2018 %>%

mutate(date = as_date(created_at, tz = "Australia/Melbourne")) %>%

count(date) %>%

ggplot(aes(date, n)) + geom_col(fill = "skyblue3") +

labs(x = "Date", y = "Tweets", title = "#user2018 tweets per day") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

Filtered for dates July 10-15, Brisbane time.

```{r tweets-by-day-hour}

user2018 %>%

mutate(datetime = as_datetime(created_at, tz = "Australia/Brisbane"), hour = hour(datetime)) %>%

group_by(date = as_date(datetime), hour) %>%

summarise(count = n()) %>%

filter(date >= as_date("2018-07-10"), date <= as_date("2018-07-15")) %>%

ggplot(aes(hour, count)) + geom_col(fill = "skyblue3") +

facet_grid(strftime(date, "%b %d") ~ .) +

labs(x = "Hour", y = "Tweets", title = "#user2018 tweets by time of day") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r tweets-top-users}

user2018 %>%

count(screen_name) %>%

filter(n >= 25) %>%

ggplot(aes(reorder(screen_name, n), n)) +

geom_col(fill = "skyblue3", color = "skyblue3") +

coord_flip() +

labs(x = "Screen Name", y = "Tweets", title = "#user2018 tweets by user", subtitle = "users with >= 25 tweets") +

theme(axis.text = element_text(size = 8),

axis.title = element_text(size = 12))

```

```{r tweets-top-sources}

user2018 %>%

distinct(screen_name, source) %>%

count(source) %>%

filter(n >= 5) %>%

ggplot(aes(reorder(source, n), n)) + geom_col(fill = "skyblue3") + coord_flip() +

labs(x = "Source", y = "Tweets", title = "#user2018 tweets by source", subtitle = "sources with >= 5 tweets") +

theme(axis.text = element_text(size = 12),

axis.title = element_text(size = 12))

```

The "replies network", composed from users who reply directly to one another.

Better to view the original PNG file in the `data` directory.

```{r reply-network, eval=FALSE}

user2018_replies <- user2018 %>%

filter(!is.na(reply_to_screen_name)) %>%

select(screen_name, reply_to_screen_name) %>%

graph.data.frame(directed = TRUE)

V(user2018_replies)$label <- V(user2018_replies)$name

V(user2018_replies)$id <- V(user2018_replies)$name

write_graph(user2018_replies, file = "../../data/user2018_replies.graphml", format = "graphml")

```


The "mentions network", where users mention other users in their tweets. Filtered for K-core >= 3.

Better to view the original PNG file in the `data` directory.

```{r mentions-network, eval=FALSE}

user2018_mentions <- user2018 %>%

filter(!is.na(mentions_screen_name)) %>%

select(screen_name, mentions_screen_name) %>%

unnest(mentions_screen_name) %>%

mutate(mentions_screen_name = strsplit(mentions_screen_name, " ")) %>%

unnest(mentions_screen_name) %>%

graph.data.frame()

V(user2018_mentions)$label <- V(user2018_mentions)$name

V(user2018_mentions)$id <- V(user2018_mentions)$name

write_graph(user2018_mentions, file = "../../data/user2018_mentions_k3.graphml", format = "graphml")

```


```{r is-retweet}

user2018 %>%

count(is_retweet) %>%

ggplot(aes(is_retweet, n)) + geom_col(fill = "skyblue3") +

labs(x = "Is retweet", y = "Tweets", title = "#user2018 tweets by retweet status") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r retweet-count}

user2018 %>%

ggplot(aes(retweet_count)) + geom_histogram(bins = max(user2018$retweet_count), fill = "skyblue3") +

labs(x = "Retweet count", y = "Tweets", title = "#user2018 distribution of retweets per tweet") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r most-retweeted}

user2018 %>%

filter(is.na(retweet_status_id)) %>%

select(screen_name, text, retweet_count) %>%

arrange(desc(retweet_count)) %>%

distinct() %>%

slice(1:10) %>%

pander(justify = c("left", "left", "right"), split.table = Inf)

```

```{r has-favorite}

user2018 %>%

mutate(has_favorite = ifelse(favorite_count > 0, TRUE, FALSE)) %>%

count(has_favorite) %>%

ggplot(aes(has_favorite, n)) + geom_col(fill = "skyblue3") +

labs(x = "Has favorite", y = "Tweets", title = "#user2018 tweets by favorited status") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r favorite-count}

user2018 %>%

ggplot(aes(favorite_count)) + geom_histogram(bins = max(user2018$favorite_count), fill = "skyblue3") +

labs(x = "Favorite count", y = "Tweets", title = "#user2018 distribution of favorites per tweet") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r most-favorited}

user2018 %>%

select(screen_name, text, favorite_count) %>%

arrange(desc(favorite_count)) %>%

distinct() %>%

slice(1:10) %>%

pander(justify = c("left", "left", "right"), split.table = Inf)

```

```{r is-quote}

user2018 %>%

count(is_quote) %>%

ggplot(aes(is_quote, n)) + geom_col(fill = "skyblue3") +

labs(x = "Is quote", y = "Tweets", title = "#user2018 tweets by quote status") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r quotes-count}

user2018 %>%

filter(!is.na(quoted_status_id)) %>%

count(quoted_status_id) %>%

ggplot(aes(n)) + geom_histogram(bins = 10, fill = "skyblue3") +

labs(x = "Quote count", y = "Tweets", title = "#user2018 distribution of quotes per tweet") +

scale_x_continuous(limits = c(0, 10), breaks = seq(0, 10, 2)) +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r most-quoted}

user2018 %>%

filter(!is.na(quoted_status_id)) %>%

count(quoted_status_id) %>%

filter(n > 1) %>%

arrange(desc(n)) %>%

inner_join(select(user2018, screen_name, quoted_status_id, is_retweet, text)) %>%

filter(is_retweet == FALSE) %>%

select(screen_name, text, quote_count = n) %>%

distinct() %>%

slice(1:10) %>%

pander(justify = c("left", "left", "right"), split.table = Inf)

```

```{r has-media}

user2018 %>%

select(media_url) %>%

unnest() %>%

mutate(has_media = ifelse(is.na(media_url), FALSE, TRUE)) %>%

count(has_media) %>%

ggplot(aes(has_media, n)) + geom_col(fill = "skyblue3") +

labs(x = "Has media", y = "Tweets", title = "#user2018 tweets by media status") +

theme(axis.text = element_text(size = 12), axis.title = element_text(size = 12))

```

```{r liked-media}

user2018 %>%

unnest(media_url) %>%

filter(!is.na(media_url)) %>%

arrange(desc(favorite_count)) %>%

select(screen_name, text, favorite_count) %>%

slice(1:10) %>%

pander(justify = c("left", "left", "right"), split.table = Inf)

```

![](`r user2018 %>%

unnest(media_url) %>%

filter(!is.na(media_url)) %>%

arrange(desc(favorite_count)) %>%

select(media_url) %>%

slice(1) %>%

unlist(use.names = FALSE)`)

The 100 words used 3 or more times.

```{r count-words}

data("stop_words")

user2018 %>%

filter(is_retweet == FALSE) %>%

unnest_tokens(word, text) %>%

select(word) %>%

filter(!word %in% c("user2018", "#user2018", "rstats", "rstat", "https", "t.co", "amp", "brisbane"),

!word %in% tolower(user2018$screen_name),

!grepl("^\\d+$", word)) %>%

anti_join(stop_words) %>%

count(word) %>%

with(wordcloud(word, n, max.words = 100, min.freq = 3, colors = brewer.pal(8, "Accent")))

```