loop through a filter call in R-CodePudding

I have a dataframe of genetic data called 'windows'

glimpse(windows)
Rows: 3,000
Columns: 5
$ chrpos       <chr> "1:104159999-104168402", "1:104159999-104168402", "1:104159999-104168402"
$ target       <chr> "AMY2A", "CFD13", "PUTA"
$ name         <chr> "rs5753", "rs70530", "rs21111"
$ chr          <chr> "1", "1", "1"
$ pos          <int> 104560629, 104562750, 104557705

I want to make them in to separate dataframes by the row in the column 'target'

AMY2A<- filter(windows, target == 'AMY2A')
write.table("am2ya.txt", header=T)

Which works fine, but is laborious doing for each element row of 'target'

But how do I loop through all of these, and then save them out as text files, in one go?

list <- windows$target
for(i in list)
df.list[[i]]<-filter(windows, target == list[[i]])

Which gives the error:

Error in `filter()`:
! Problem while computing `..1 = target == list[[i]]`.
Caused by error in `list[[i]]`:
! subscript out of bounds

And when I google saving out multiple txt files, it just comes up how to read in multiple text files.

Any help would be great, thanks!

CodePudding user response：

Or, using the tidyverse, something like

windows %>% 
  group_by(target) %>% 
  group_walk(
    function(.x, .y) {
      write.table(.x, paste0(tolower(.y$target[1]), ".txt"))
    },
    .keep=TRUE
  )

[Untested code, since your question is not reproducible.]

the .keep=TRUE keeps the target column in the text file. If that's unnecessary, feel free to delete it. Details are in the online doc.

CodePudding user response：

probably something like:

lst <- windows$target

for(i in seq_along(lst)){
  dat <- filter(windows, target == lst[[i]])
  write.table(dat, paste0(tolower(lst[[i]]), ".txt"))
}