From the list of data frame select particular columns in R

Question

I have multiple CSV files say 2, and I want to extract the specific column from each file.

filenames <- list.files("path")
for(iter in 1:length(filenames)){
  print(filenames[iter])
  my_csv=read.csv(filenames[iter])

Where, csv1 and csv2 contains columns a1,a2,a3,a4 and b1,b2,b3,b4 respectively. Now I want to fetch column a1 from csv1 and b2 from csv2 where csv number (csv1(1 in this case) matches with coulmn name (a1 in this case)). Likewise b2 from csv2(csv2 matches with colummn name b2.

So from file csv1.csv you want to extract first column, or column that has "1" in it? — pogibas, Aug 18 '17 at 03:42
@PoGibas, I want to extract a column "a.filename" (in this case a.csv1). — Kaushik, Aug 18 '17 at 03:49
Files are named as csv1.csv, csv2.csv. And every csv contains around 100 columns. But I want to fetch only those columns named as a.csv1(column name in csv1 file), a.csv2(column name in csv2 file). a is common in every file. I have 3 files in total. Final output of columns should be a.csv1,a.csv2,a.csv3. — Kaushik, Aug 18 '17 at 03:58

pogibas · Accepted Answer · 2017-08-18T04:09:47.027

0

For reading in csv files I strongly recommend using fread from data.table package (More on that here).

library(data.table)
fileCSV <- list.files(pattern = "csv$")
# This function extracts i column from the csv file
# It doesn't look into colnames
getCSVcolumn <- function(FILE) {
    i <- gsub("csv|\\.", "", FILE)
    foo <- fread(FILE)
    foo[, paste0("V", i), with = FALSE]
}
res <- sapply(fileCSV, getCSVcolumn)
res <- do.call("cbind", res)
colnames(res) <- gsub("\\..*","", colnames(res))

edited Aug 18 '17 at 04:09

answered Aug 18 '17 at 03:39

pogibas

27,303
19
84
117

Thanks @PoGibas. It is working. But I am getting all the columns in one data frame. I want separate data frame for each column. – Kaushik Aug 18 '17 at 04:21
Just don't use anything after apply part – pogibas Aug 18 '17 at 04:43

From the list of data frame select particular columns in R

1 Answers1