Twitter_weekly_download

### Workflow for weekly download of twitter data with specific hashtags (keywords)###

### The code was developed as part of the IARH project, in order to download all the tweets containg specifc hashtag the appear every week.
### The code downloads only the tweets published up to 7-10 ago from the time it is run - hence will be re-run every week 
### Project IARH
### Author Marta Krzyzanska

#Set working direcotry:

setwd("~path")

# Require rtweet. If you have not installed it already, run install.packages() first.

library("rtweet")

# Load or create token (for the project the twitter_token is located on the google drive/R useful files/twitter):

load("twitter_token.R")

# Load or create list of hashtags to search for:

load("twitter.R")

#Create a list in which to store tweets with different hashtags

tweets <- c()

#Loop through all the hashtags/keywords, extract the relevant tweets. Wait 15 minutes between each hashtag/keywords, because twitter API
#has a limit on API calls within 15 minutes and usually it is used up in one search
#It will need to be edited if any of the hashtags returns 18000 tweets, which will suggest that there are more 
#(search_tweets returns maximum of 18000 tweets, which should be sufficient for our hashtag, otherwise we may need to use alternative 
# method which is slower and sometimes return errors)

i=1
j=length(twitter_keywords)+1

while (i<j){

tweets[[i]]<-search_tweets(q=twitter_keywords[i], n = 18000, token=twitter_token)
print (paste("Search for tweets containing keyword: ",twitter_keywords[i]," completed",sep=""))
print ("Waiting for the token reset")
Sys.sleep(900)
i=i+1}

save(tweets,file="tweets(#enter_download_date).R")