CODEWITHSUNDEEP
r
Cecile
Cecile

Reputation: 535

creating a series of files duplicates and renaming them

I have a list of climate files in a folder (EDU link: https://cloudstor.aarnet.edu.au/plus/index.php/s/QFpaBDR7q2GbDCN). They are named as follow p[year][month]_0p5.asc.

Example:

p198001_0p5.asc

p198002_0p5.asc

...

p198012_0p5.asc

p198101_0p5.asc

...

p201012_0p5.asc

I wish to create 2000 years of data by repeating in that order these 31 years (1980 to 2010) and renaming them to have a list of files from p198001_0p5.asc to p397912_0p5.asc

As example, the final folder will include p201101_0p5.asc which contains the same data as p198001_0p5.asc but will be named p201101_0p5.asc.

I have no idea how to do this in R. Any help would be much appreciated!

Update 19-07-2017: @Mako212 suggestion seems to work but I had memory issues (error message # protect(): protection stack overflow). I changed my strategy and created another script based on his kind suggestion to generate 131 years of data from 1880 to 2010. I did this by repeating the series of 31 years (1980-2010). Here is the code in case you'd like to have a look at that strategy too. It works fine:

require(purrr)
require(data.table) 
library(raster)

setwd('...')
files <- list.files(pattern= "*.asc")
files
length(files)

fileDF <- files %>% map(raster)

#set the output directory
output_dir <- "..."

#  set month and year counters
startYear <- 2010 
startMonth <- 12
fileNumber <- 1

for (i in fileDF){
  startYear <- 2010 - (fileNumber-1) %/% 12
  for (x in 1:6){
    print(startYear)
    print(startMonth)
    print(fileNumber)
    writeRaster(i, paste(output_dir, sprintf("p%d%s%d_0p5.asc", startYear, ifelse(startMonth<10,0,""), startMonth), sep="/"), format = "ascii")
    startYear <- startYear - 31
    if (startYear < 1880) break #  don't create files before December, 1880
  }
  if (startMonth > 1 ) {
    startMonth <- startMonth - 1
  }
  else{
    startMonth <- 12
  }
  fileNumber <- fileNumber + 1    
}

Upvotes: 1

Views: 102

Answers (1)

Mako212
Mako212

Reputation: 7312

Okay, here's the general idea:

require(purrr)
require(data.table) 
#  after playing with SDMTools::read.asc, data.table::fread seems to
#  be more reliable. That said, if fread() isn't reading your data
#  correctly, you might try using the SDMTools function instead. 
#  I also chose to save everything as .csv, but again, you can try 
#  using the SDMTools read/write.asc functions if you want

files <- list.files(pattern= "*.asc")


fileDF <- files %>% map(fread)

#  set month and year counters

startYear <- 1980 
startMonth <- 1
fileNumber <- 1

for (i in fileDF){


    #  increment startYear by 1 every 13th file   
    startYear <- 1980 + (fileNumber-1) %/% 12

    for (x in 1:65){
        #  added underscore for clarity between year and month
        #  format is p1980_01_0p5.csv
        write.csv(i, sprintf("p%d_%s%d_0p5.csv", startYear, ifelse(startMonth<10,0,""), startMonth))
        startYear <- startYear + 31

        #  don't create files past December, 3979
        if (startYear > 3979) break
    }
    if (startMonth < 12) {
        startMonth <- startMonth +1
    }
    else{
        startMonth <- 1
    }

    fileNumber <- fileNumber + 1    


}

The loop counters are set assuming you have exactly 31 years worth of data (12*31 files)

Make sure you write to a new folder (not the one that contains the source data)

Upvotes: 1

Related Questions