Using forloop to add headers to multiple files

Question

[jordi.planells]

Hi guys! I need help in setting a for loop in R (I'm quite new programming in R). I would like to add the same header to all the files that match a concrete pattern inside a folder.

To get the list of files I'm using the following code:

filelist <- list.files(pattern = "DESeq2_result*")

And this is the for loop I am trying to implement:

for (i in seq_along(filelist)) {
   names[[i]] <- a
   out [i]

}

where a is a vector that I defined with the names of the different columns:

a <- c("gene_id", "baseMean", "log2FC",
       "SD", "WaldStatistic", "pval", "padj")

If you have any tutorial/page to help me to learn and practise my ability to code functions and loops in R would be so much appreciated.

Thank you so much in advance!

Jordi

[Benn]

What error do you get? What is it that you would like to get out of the loop? The same files but now with change of header?

Answer 1

[Gautier Richard]

As I understood, you want to open some files containing the pattern "DESeq2_result*" as data.frames in a list, assigning new names to the columns of these dataframes based on the a vector, and maybe name the elements of the list based on the files name? You can do it that way:

filelist <- list.files(pattern = "DESeq2_result")

a <- c("gene_id", "baseMean", "log2FC",
       "SD", "WaldStatistic", "pval", "padj")

data<-list()

for (i in 1:length(filelist)) {
  data[[i]]<-read.table(filelist[i])
  colnames(data[[i]])<-a
}

names(data)<-filelist

for (i in 1:length(data)) {
    output_name <- paste0("colnames_",names(data[i]))
    write.table(data[[i]], output_name, quote = F, row.names = F, col.names = T, sep="\t")
}

EDIT: added the write.table part to take your comment into account in order to export the imported files with the column names.

[jordi.planells]

Basically, what i want to do is to add column names to my files, in one piece of code. For a single file, the code would be:

data <- read.table("file", sep = "\t", header = FALSE)
names(data)
colnames(data) <- a
write.table(data, "file", sep = "\t", col.names = TRUE, row.names = FALSE)

I'm trying to optimise my code, so I can do all of them without having to type in my Rmarkdown document 8 times the same code (one for each file)

[Gautier Richard]

Ok, I modified the code so it does what you want.

I think that a much better way would be to simply replace the first line of every file with a shell script or a one-liner. You don't need R for that. The code above is really heavy for what you need. Pasting the first line stored in file a at the beginning of every file with pattern would be much more efficient, but not R-based.

[jordi.planells]

Ok thank you so much! So basically you are suggesting me to do this with a shell for loop and sed/awk command instead. Thank you so much again for your help! Cheers!

Answer 2

[zx8754]

Use col.names = argument when reading the files, then write out, something like this:

for(i in list.files(pattern = "DESeq2_result_*"))
  write.table(read.table(i, col.names = c("gene_id", "baseMean", "log2FC",
                                          "SD", "WaldStatistic", "pval", "padj")), i)

Note: this overwrites existing files, to create new files:

for(i in list.files(pattern = "DESeq2_result_*"))
  write.table(read.table(i, col.names = c("gene_id", "baseMean", "log2FC",
                                          "SD", "WaldStatistic", "pval", "padj")),
              paste0(i, ".fixed.txt"))

[jordi.planells]

Thank you a lot!! It is really appreciated

Answer 3

[Arup Ghosh]

Creates a new file with headers. Change the gsub pattern before using.

col=c("gene_id", "baseMean", "log2FC","SD", "WaldStatistic", "pval", "padj")
for(file in list.files(pattern="^DESeq2_result*")){
  write.table(read.table(file,col.names = col),
              paste0("updated_",gsub(pattern = "\\.txt$", "", file),".tsv"),col.names = TRUE, row.names = FALSE,sep="\t") #Also update the file extension
}

[jordi.planells]

Thank you too, arup!