How do I write a loop to manipulate dataframes in R?

I am trying to make a large table of results from a large set of dataframes in R. I have over 1000 tables I am trying to combine and so a loop or at least function is necessary to do this efficiently but I can't seem to get it working properly. I'm not sure if using dplyr::adply would be more efficient, but I'm less familiar with that type of scripting.

The tables must be imported into R, cut down to two columns, filtered, and then the resulting column needs to be appended to a growing table. The outcome would be a table with 1000 columns, one column from each data frame I am importing.

This is my manual adjustment to the file that I am trying to put into loop/function form:
<pre> <code>

#import files
x <- read.delim("~/maf_files/patient1.maf", comment.char="#")
#cut down to three relevant columns
df <- `x`[,c(1,9,16)]
#filter for Missense Mutations only
df2 <- df[df$Variant_Classification == "Missense_Mutation",]

library(tidyverse)

#filter for unique gene names
df3<-df2 %>% 
  group_by(Hugo_Symbol) %>% 
  filter(n() == 1)

#change column names
colnames(df3)[3] <- "patient1"
#remove Missense Column
df3$Variant_Classification <- NULL
#change column values to 1
df3$`patient1` <-1

#append column to existing table - should have 1000 columns
test <- full_join(test, df3, by = "Hugo_Symbol")</kbd>

This is what I tried to do first, making a function that uses all of these steps:

<pre> <code>

read_maf <- function(file){
x <- read.delim("~/maf_files/file.maf", comment.char="#")
df <- x[,c(1,9,16)]
df2 <- df[df$Variant_Classification == "Missense_Mutation",]
df3<-df2 %>% 
group_by(Hugo_Symbol) %>% 
filter(n() == 1)

colnames(df3)[3] <- "patient1"
df3$Variant_Classification <- NULL
df3$`patient1` <-1
test <- full_join(test, df3, by = "Hugo_Symbol")

})

try something like what is below. that is your function that reads and adjust each individal files. to bind them all into a single dataframe, you can use purrr::map_dfr. list_of_files should be a list of all the files you are reading in and want to bind. like so:

list_of_files <- ('path1.maf','path2.maf')

read_maf <- function(file){
x <- read.delim(glue("~/maf_files/{file}.maf"), comment.char="#")
df <- x[,c(1,9,16)]
df2 <- df[df$Variant_Classification == "Missense_Mutation",]
df3<-df2 %>% 
group_by(Hugo_Symbol) %>% 
filter(n() == 1)

colnames(df3)[3] <- "patient1"
df3$Variant_Classification <- NULL
df3$patient1 <-1
return(df3})

big_frame <- purrr::map_dfr(list_of_files,read_maf)

Multiple things are going on here. First, your function will always return the same because you didn't indicate where the file element should be used. Assuming it refers to the file location, you should change read.delim("~/maf_files/file.maf", comment.char="#") with read.delim(file, comment.char="#") inside the function. Also, I would add return(test) at the end of the function. Ensure it works properly with read_maf("~/maf_files/file.maf") . It should return the same data frame as in your first code.

Second, you need a vector with the location of each one of the files that you are going to loop. You can get it with the following code.

files <– list.files("~/maf_files", ".maf", full.names = T)

Once you have the function working properly and the files' location, you should loop the files’ location with your created function. Since you want to join everything in one data frame, I recommend you use purrr:map_dfr(). It does the same than lapply() but joins everything at the end. Something like this:

final_df <- purrr:map_dfr(files, read_maf)