问题:

I have a .csv file with data for different chromosomes. The chromosomes names are stored in the first column(column name: Chr). My aim is to separate the data for each chromosome i.e. (Chr1,Chr2 etc) and make separate csv files for each. I cannot understand how to do this in limited steps. Thanks

回答1:

Illustrating a one liner using plyr and the dataset iris

plyr::d_ply(iris, .(Species), function(x) write.csv(x, 
  file = paste(x$Species, ".csv", sep = "")))

回答2:

Read Data

fn <- dir(pattern="csv")
data_in <- do.call(rbind,lapply(fn,read.csv))

Split by chromosome

data_out <- split(data_in,data_in[[1]])

Write by chromosome

chn <- unlist(lapply(data_out,"[",1,1))
for(i in seq_along(chn)) write.csv(data_out[[i]],file=paste(chn[i],"csv",sep="."))

回答3:

One way is to read the input file one line at a time and append the line to the correct outfile based on the first x characters of the line:

con <- file('yourInputFile', 'r') 

while (length(input <- readLines(con, n=1) > 0){ 
      outputfile <- paste(substr(input, 1, 5), ".csv", sep="" ) 
       ### assuming first 5 characters are new file name
      outfile <- file(outputfile, 'a')
      writeLines(output, con=outfile)
      close(outfile)
}

The advantage of this approach is that it works even if yourInputFile is too big to read into memory. The downside is that this approach is slow as it does a lot of opening/closing of files.