Convert Stata 16 files to Stata 12 files using R

问题

I am using RStudio (running R 4.0.1) and Stata 12 for Windows and have got a large number of folders with Stata 16 .dta files (and other types of files not relevant to this question). I want to create an automated process of converting all Stata 16 .dta files into Stata 12 format (keeping all labels) to then analyze. Ideally, I want to keep the names of the original folders and files but save the converted versions into a new location.

This is what I have got so far:

setwd("C:/FilesLocation")
#vector with name of files to be converted
all_files <- list.files(pattern="*.dta",full.names = TRUE)
for (i in all_files){

#Load file to be converted into STATA12 version
data <- read_dta("filename.dta",
                 encoding = NULL,
                 col_select = NULL,
                 skip = 0,
                 n_max = Inf,
                 .name_repair = "unique")

#Write as .dta
write_dta(data,"c:/directory/filename.dta", version = 12, label = attr(data, "label"))
}

Not sure this is the best approach. I know the commands inside the loop are working for a single file but not really being able to automate for all files.

回答1:

Your code only needs some very minor modifications. I've indicated the changes (along with comments explaining them) in the snippet below.

library(haven)

mypath <- "C:/FilesLocation"
all_files <- list.files(path = mypath, pattern = "*.dta", full.names = TRUE)

for (i in 1:length(all_files)){ 
  #(Above) iterations need the length of the vector to be specified

  #Load file to be converted into STATA12 version
  data <- read_dta(all_files[i], #You want to read the ith element in all_files
                   encoding = NULL,
                   col_select = NULL,
                   skip = 0,
                   n_max = Inf,
                   .name_repair = "unique")

  #Add a _v12 to the filename to 
  #specify that is is version 12 now
  new_fname <- paste0(unlist(strsplit(basename(all_files[i]), "\\."))[1], 
                     "_v12.", unlist(strsplit(basename(all_files[i]), "\\."))[2])

  #Write as .dta
  #with this new filename
  write_dta(data, path = paste0(mypath, "/", new_fname), 
            version = 12, label = attr(data, "label"))

}

I tried this out with some .sta files from here, and the script ran without throwing up errors. I haven't tested this on Windows but in theory it should work fine.

Edit: here is a more complete solution with read_dta and write_dta wrapped into a single function dtavconv. This function also allows the user to convert version numbers to arbitrary values (default is 12).

#----
#.dta file version conversion function
dtavconv <- function(mypath = NULL, myfile = NULL, myver = 12){

  #Function to convert .dta file versions
  #Default version files are converted to is v12
  #Default directory is whatever is specified by getwd()

  if(is.null(mypath)) mypath <- getwd()

  #Main code block wrapped in a tryCatch()
  myres <- tryCatch(
    {

      #Load file to be converted into STATA12 version
      data <- haven::read_dta(paste0(mypath, "/", myfile),
                              encoding = NULL,
                              col_select = NULL,
                              skip = 0,
                              n_max = Inf,
                              .name_repair = "unique")

      #Add a _v12 to the filename to 
      #specify that is is version 12 now
      new_fname <- paste0(unlist(strsplit(basename(myfile), "\\."))[1], 
                          "_v", myver, ".", unlist(strsplit(basename(myfile), "\\."))[2])

      #Write as .dta
      #with this new filename
      haven::write_dta(data, path = paste0(mypath, "/", new_fname), 
                       version = myver, label = attr(data, "label"))

      message("\nSuccessfully converted ", myfile, " to ", new_fname, "\n")

    }, 
    error = function(cond){

      #message("Unable to write file", myfile, " as ", new_fname)
      message("\n", cond, "\n")
      return(NA)

    }
  )

  return(myres)
}
#----

The function can then be run on as many files as desired by invoking it via lapply or a for loop, as the example below illustrates:

#----

#Example run
library(haven)

#Set your path here below
mypath <- paste0(getwd(), "/", "dta")

#Check to see if this directory exists
#if not, create it
if(!dir.exists(mypath)) dir.create(mypath)
list.files(mypath)
# character(0)

#----
#Downloading some valid example files
myurl <- c("http://www.principlesofeconometrics.com/stata/airline.dta", 
           "http://www.principlesofeconometrics.com/stata/cola.dta")
lapply(myurl, function(x){ download.file (url = x, destfile = paste0(mypath, "/", basename(x)))})

#Also creating a negative test case
file.create(paste0(mypath, "/", "anegcase.dta"))

list.files(mypath)
# [1] "airline.dta"  "anegcase.dta" "cola.dta" 
#----


#Getting list of files in the directory
all_files <- list.files(path = mypath, pattern = "*.dta")


#Converting files using dtavconv via lapply
res <- lapply(all_files, dtavconv, mypath = mypath)
# 
# Successfully converted airline.dta to airline_v12.dta
# 
# 
# Error in df_parse_dta_file(spec, encoding, cols_skip, n_max, skip, 
# name_repair = .name_repair): Failed to parse /my/path/
# /dta/anegcase.dta: Unable to read from file.
# 
# 
# 
# Successfully converted cola.dta to cola_v12.dta
# 

list.files(mypath)
# [1] "airline_v12.dta" "airline.dta"     "anegcase.dta"    "cola_v12.dta"    
# "cola.dta" 

#Example for converting to version 14
res <- lapply(all_files, dtavconv, mypath = mypath, myver = 14)
# 
# Successfully converted airline.dta to airline_v14.dta
# 
# 
# Error in df_parse_dta_file(spec, encoding, cols_skip, n_max, skip, 
# name_repair = .name_repair): Failed to parse /my/path
# /dta/anegcase.dta: Unable to read from file.
# 
# 
# 
# Successfully converted cola.dta to cola_v14.dta
# 

list.files(mypath)
# [1] "airline_v12.dta" "airline_v14.dta" "airline.dta"     "anegcase.dta"    
# "cola_v12.dta"    "cola_v14.dta"    "cola.dta" 

#----

来源：https://stackoverflow.com/questions/62399350/convert-stata-16-files-to-stata-12-files-using-r

标签

stata