arrange_gc() Function

This function creates table from selected columns of report files generated through Agilent GC ChemStation. In order to prepare files for loading, user should put all folders that contain REPORT01.CSV (default name) into one folder. Function ‘arrange_gc’ search for ‘REPORT01.CSV’ files in each sub-folder, takes values of retention time (RT-FID) and area from it, calculate each peak share in percent (m/m%) and arrange that two columns in new table (final.csv). Arguments: directory - path to initial folder max - maximal count of peaks. This basically depend on ‘Integration events’ sol - default value is FALSE, but if first peak is peak of solvent user should put ‘sol = TRUE’ new_soft - since Agilent ChemStations produce different CSV files depending software version. User should put ‘new_soft = FALSE’ if software is older than 2003.

arrange_gc <- function(directory, max = 100, sol = FALSE, new_soft = TRUE) {
  encoding <- c("UTF-16") # set default encoding value
  if(new_soft != TRUE){encoding <- c("UTF-8")} # set encoding value for old software
  fld <- list.dirs(directory) # get list of sub-folders together with parent folder
  fld_count <- length(fld) # get number of sub-folders together with parent folder
  fld <- fld[2:fld_count] # exclude parent folder from a vector
  fld_names <- basename(fld) # create vector of sub-folder names
  final_df <- data.frame(1:max) # create one-column data frame with 'max' number of rows
  count <- length(fld_names) # get count of sub-folders only
  
  for(i in 1:count){
    df <- read.table(file.path(directory, fld_names[i], "REPORT01.csv"), sep =",", 
                     header = FALSE, fileEncoding = encoding) # read REPORT01.CSV from each sub-folder
    if(sol == TRUE){df <- df[-c(1),]} # remove solvent peak area if exists 
    df[, 2] = round(df[,2],3) # set retention time on two decimals
    area_sum <- sum(df[, 5]) # get sum value of area column
    mm <- round(df[, 5] * 100 / area_sum, 2) # calculate peak share in percent, round value on two decimals 
    new_df <- data.frame(df[, 2], mm) # create new data frame with chosen columns - retention time and percentage
    rows_count <- nrow(new_df) # get row number of current data set
    diff <- max - (rows_count + 1) # calculate difference between current data row numbers and predicted 'max' value
    new_df <- rbind(c(fld_names[i], fld_names[i]), new_df) # create header with sub-folder names
    for(i in 1:diff){
      new_df <- rbind(new_df, c(NA, NA)) # fill rest of the rows in data set with NA's
    }
    colnames(new_df) <- c("RT-FID", "%m/m") # set column names
    final_df <- cbind(final_df, new_df, row.names = NULL) # create new data frame for final table
  }
  final_df <- final_df[,-c(1)] # delete initial column of data frame
  write.csv(final_df, file = (file.path(directory, "final.csv"))) # write data frame to file
  print("File 'final.csv' has been created in your initial directory.") # return information to user
}

Let’s try out this function. Note that argument for software version is set to FALSE.

arrange_gc("C:/Users/dejan/Documents/R/GC folder", new_soft = FALSE)

## [1] "File 'final.csv' has been created in your initial directory."

This is a part of the output file ‘final.csv’.

##            V2         V3         V4         V5         V6
## 1      RT-FID       %m/m   RT-FID.1     %m/m.1   RT-FID.2
## 2  GFID3891.D GFID3891.D GFID3892.D GFID3892.D GFID3893.D
## 3        9.07       0.05      9.297       0.04      9.408
## 4       9.349       0.04      9.427       0.58      10.12
## 5        9.48       0.72     10.142       0.91     10.259
## 6      10.195       0.78     10.281       0.51     12.416
## 7      10.334       0.42      12.44       0.11     13.643
## 8      12.496       0.08     13.669       0.05     14.311
## 9      13.723       0.06     16.223       0.13     16.197
## 10     14.392       0.03     17.171       0.09     16.611