Every 5 minutes get the google trend data

Question

I would like to extract the google trend data through package "gtrendsR" every hour, I tried to use Sys.sleep() function to set the timer, however, it is failure for me to download it every hour. So, how do I correct my code in order to get the data every hour. Thanks a lot!

Sys.setlocale("LC_ALL", "English")
Keywords = c("google", "twitter")

for (k in Keywords) {
  res = NULL

  temp_1 <- gtrends(k, geo = "US",time = "all")
  temp_2 <- temp_1$interest_over_time
  res <- rbind(res, temp_2)
  rm(temp_1,temp_2)

  res <- select (res, c(date, hits))

  Sys.setlocale(category = "LC_ALL", locale = "cht")
  names(res)[2]<- k
  xfilepath = paste("C:/Users/Peter/Desktop/",k,".csv",sep="")
  write.csv(res, file = xfilepath, row.names=FALSE)

}

Answer 1

Sys.sleep() should be working, and without an error code it is hard to tell you exactly why it is failing. I'll suggest an alternative method however.

The later package is a simple and nice utility package for executing code, well.. Later. It takes a function without any arguments, and runs it at a set delay. For example you could use:

library("gtrendsR")
library("later")
library("data.table") #for as.ITime
Sys.setlocale("LC_ALL", "English")
Keywords = c("google", "twitter")
#Set delay. Here for 5 seconds
delay <- as.ITime("00:00:05")
Interrupt <- FALSE

extractGoogle <- function(){
    for (k in Keywords) {
        res = NULL
        temp_1 <- gtrends(k, geo = "US",time = "all")
        temp_2 <- temp_1$interest_over_time
        res <- rbind(res, temp_2)
        rm(temp_1,temp_2)
        res <- select (res, c(date, hits))
        Sys.setlocale(category = "LC_ALL", locale = "cht")
        names(res)[2]<- k
        xfilepath = paste("C:/Users/Peter/Desktop/",k,".csv",sep="")
        write.csv(res, file = xfilepath, row.names=FALSE)
    }
    #Execute once again later
    if(isFALSE(Interrupt))
        later(extractGoogle, delay = delay)
}
#Run the function until Interrupt is set to TRUE or session closes
extractGoogle()

This allows you to set the delay manually, changing 'delay' to a number of seconds. as.ITime simple lets you specify the number of seconds in a simple format. The cycle can then be delayed further or broken by changing the global variables.