Código para captura de busca de tuítes

analise_popularidade.R

library(lubridate)
library(scales)
library(tidyverse)

# regulariza o formato da data do tweet, corrige para fuso certo
d$created_at <- as.POSIXct(strptime(d$created_at, "%Y-%m-%d %H:%M:%S"))
d$created_at <- d$created_at - hours(3)
d$dia <- as.POSIXct(strptime(d$created_at, "%Y-%m-%d"))

# cria coluna com minuto arredondado
# alternativas de agregação “secs”, “mins”, “hours”, “days”, “months”, “years”
d$min_redondo <- round(d$created_at, units = "hours")
d$new <- as.character(d$min_redondo)

# transformat status_id em caractere, em vez de numero
d$status_id <- as.character(d$status_id)

contagem <- d %>%
  group_by(new) %>%
  #group_by(dia) %>%
  #summarize(count=n())
  count()

contagem$new <- as.POSIXct(strptime(contagem$new, "%Y-%m-%d %H:%M:%S"))
contagem$new <- contagem$new - hours(3)
#filtrado <- contagem %>%
#filter(new > "2019-11-07 23:59:00")
ggplot(contagem, aes(new, n)) + 
  geom_point(alpha = 0.6) + 
  geom_bar(stat="identity") 

get_tweets.R

library(rtweet)

## armazena as chaves do Twitter
api_key <- ""
api_secret_key <- ""
access_token <- "0"
access_token_secret <- ""

## faz autenticação via navegador
token <- create_token(
  app = "popularity_R",
  consumer_key = api_key,
  consumer_secret = api_secret_key,
  access_token = access_token,
  access_secret = access_token_secret)

# faz a busca 
tt_corona <- search_tweets(
  q = 'coronavirus OR covid-19 -filter:retweets -filter:replies',
  n = 18000,
  type = "recent",
  include_rts = TRUE,
  geocode = NULL,
  max_id = '',
  parse = TRUE,
  token = NULL,
  lang = "pt",
  retryonratelimit = TRUE,
  verbose = TRUE)

# cria nofo DF
d <- tt_corona

# faz um unlist das colunas que são listas, para salvar csv
d0 <- subset(d, select=-c(symbols,
                          urls_expanded_url,
                          media_t.co,
                          media_expanded_url,
                          ext_media_url,
                          ext_media_t.co,
                          ext_media_expanded_url,
                          geo_coords,
                          coords_coords,
                          bbox_coords,
                          urls_t.co
))

# limpeza de colunas
d0$hashtags <- vapply(d0$hashtags, paste, collapse = ", ", character(1L))
d0$mentions_user_id <- vapply(d0$mentions_user_id, paste, collapse = ", ", character(1L))
d0$mentions_screen_name <- vapply(d0$mentions_screen_name, paste, collapse = ", ", character(1L))
d0$urls_url <- vapply(d0$urls_url, paste, collapse = ", ", character(1L))
d0$media_url <- vapply(d0$media_url, paste, collapse = ", ", character(1L))
d0$media_type <- vapply(d0$media_type, paste, collapse = ", ", character(1L))

# escreve o CSV
write.csv(d0, "tweets_covid_XX_XX_marco.csv")