#This script should be run on 1 node for 8 hours on the cluster

#The organization script loads in and filters the AR data into 12 large data frames (one for each month) 
#with the frequency of AR hits by location. 

library(dplyr)
library(lubridate)

#setwd("Y:/hwalcek/ar_2010_all")
setwd("/depot/wwtung/data/hwalcek/ar_2010_all")

#combine all csv files - csv files are in the format ar_number, so a loop is necessary
for (Number in 1:1){
  filename <- paste("ar_",Number,".csv", sep = "") #load in first csv file
  ar_all <- read.csv(filename, sep=",")
}

for (Number in 2:2920){
  filename <- paste("ar_",Number,".csv", sep = "") #load in all csv files
  ar_all_data <- read.csv(filename, sep=",") #read csv files
  ar_all <- rbind(ar_all, ar_all_data) #combine with previously read in csv file
}


#save(ar_all, file="Y:/hwalcek/ar_2010_monthly/ar_all.Rdata")
save(ar_all, file = "/depot/wwtung/data/hwalcek/ar_2010_monthly/ar_all.Rdata")

#load("Y:/hwalcek/ar_2010_monthly/ar_all.Rdata") 
load("/depot/wwtung/data/hwalcek/ar_2010_monthly/ar_all.Rdata")
  
#add separate year, month, day columns
ar_all_sep <- ar_all %>% 
dplyr::mutate(year = lubridate::year(Time), 
              month = lubridate::month(Time), 
              day = lubridate::day(Time))

#divide dataframe by month
ar_monthly <- group_split(ar_all_sep, ar_all_sep$month) 



#Frequency of AR hits by location each month:
  
#setwd("Y:/hwalcek/ar_2010_monthly")
setwd("/depot/wwtung/data/hwalcek/ar_2010_monthly")
  
#add AR counts together to create frequency
for (Number in 1:12){
  month_counts <- ar_monthly[[Number]] %>%
    group_by(Lon, Lat)%>%
    transmute(count = sum(AR))
    filename <- paste("ar_month_",Number,".rds", sep = "") #create file name
    saveRDS(month_counts, file = filename)} #save each file to an rds