R/load_file.r

In roqua/autovar: Automated Vector Autoregression Model Creation

#' Load a data set from a .sav, .dta, or .csv file
#'
#' This function prints the columns of the loaded data set. The abbreviation \code{(scl)} is used to denote scale (numeric) columns, and \code{(nom)} is used to denote nominal (factor) columns. The function returns an object of class \code{av_state}. Objects of this class are used throughout this package to store VAR data sets, models, and results.
#' @param filename path and filename of the data set to load. The path can be relative to the current working directory.
#' @param file_type specifies whether the file contains STATA, SPSS, or CSV data. This argument is optional. When not specified, it is determined from the file name, i.e., .dta extensions are treated as STATA files, .sav extensions are treated as SPSS files, and .csv extensions are treated as CSV files.
#' @param log_level sets the minimum level of output that should be shown (a number between 0 and 3). A lower level means more verbosity. Specify a log_level of 3 to hide messages about the data set that was loaded.
#' @return This function returns the newly created \code{av_state} object.
#' @examples
#' \dontrun{
#' av_state <- load_file("../data/input/RuwedataAngela.sav")
#' print(av_state)
#' }
#' @export
load_file <- function(filename,file_type = c('SPSS','STATA','CSV'),log_level=0) {
  av_state <- new_av_state()
  real_file_name <- filename
  if (!is.null(attr(filename,"filename"))) {
    real_file_name <- attr(filename,"filename")
  }
  if (missing(file_type)) {
    file_type <- determine_file_type(real_file_name)
  }
  working_dir <- paste(getwd(),'/',sep="")
  file_name <- paste(working_dir,filename,sep="")
  if (!file.exists(file_name)) {
    if (!file.exists(filename)) {
      stop(paste("File does not exist:",filename))
    }
    # NOTE: filename needs to include the full and valid path for SPSS export to work
    # if the data is stored in a separate file
    file_name <- filename
  }
  av_state$file_name <- file_name
  av_state$real_file_name <- real_file_name
  av_state$file_type <- match.arg(file_type)

  av_state <- switch(av_state$file_type,
    SPSS = load_file_spss(av_state),
    STATA = load_file_stata(av_state),
    CSV = load_file_csv(av_state)
  )
  scat(log_level,2,"load_file loaded",av_state$file_type,"file with",length(av_state$data[[1]]),"columns:\n")
  sprint(log_level,2,column_names_output(av_state))
  av_state
}

column_names_output <- function(av_state) {
  dnames <- NULL
  for (name in names(av_state$data[[1]])) {
    column_type <- decorate_class(av_state$data[[1]][[name]])
    dnames <- c(dnames,paste(name," (",column_type,")",sep=""))
  }
  dnames
}

decorate_class <- function(data) {
  if (is(data, "factor")) {
    "nom"
  } else if (is(data, "numeric")) {
    "scl"
  } else if (is(data, "integer")) {
    "scl"
  } else {
    NULL
  }
}

determine_file_type <- function(filename) {
  fileonly <- basename(filename)
  if (length(grep("\\.dta$",fileonly)) == 1) {
    'STATA'
  } else if (length(grep("\\.sav$",fileonly)) == 1) {
    'SPSS'
  } else if (length(grep("\\.csv$",fileonly)) == 1) {
    'CSV'
  } else {
    warning(paste("Unrecognized file type:",filename))
  }
}

load_file_spss <- function(av_state) {
  av_state$raw_data <- read_spss(av_state$file_name, to.data.frame=TRUE)
  av_state$data <- list(multiple=av_state$raw_data)
  av_state
}

load_file_stata <- function(av_state) {
  av_state$raw_data <- foreign::read.dta(av_state$file_name)
  av_state$data <- list(multiple=av_state$raw_data)
  av_state
}

load_file_csv <- function(av_state) {
  # try to determine the separator
  fd <- file(av_state$file_name)
  open(fd)
  first_line <- readLines(fd,n=1)
  close(fd)
  if (stringr::str_count(first_line,";") >= stringr::str_count(first_line,","))
    av_state$raw_data <- read.csv2(av_state$file_name)
  else
    av_state$raw_data <- read.csv(av_state$file_name)
  av_state$data <- list(multiple=av_state$raw_data)
  av_state
}