R/validation_check.R

In triplediff: Triple-Difference Estimators

# This process run all the error check before enter to the data processing

#' @importFrom Rcpp sourceCpp

validate_args_2Periods <- function(args, dta){
  # get args
  yname <- args$yname
  tname <- args$tname
  idname <- args$idname
  gname <- args$gname
  pname <- args$pname
  xformla <- args$xformla
  est_method <- args$est_method
  base_period <- args$base_period
  learners <- args$learners
  n_folds <- args$n_folds
  weightsname <- args$weightsname
  boot <- args$boot
  nboot <- args$nboot
  inffunc <- args$inffunc
  cband <- args$cband

  # flag for boot and cband
  if ((!boot) && (cband)){
    stop("cband is only available when boot = TRUE")
  }

  # Flag for yname
  if (!is.element(yname, base::colnames(dta))) {
    stop("yname = ",yname,  " could not be found in the data provided.")
  }

  # Flag for tname
  if (!is.element(tname, base::colnames(dta))) {
    stop("tname = ",tname,  " could not be found in the data provided.")
  }

  # check if times periods are numeric
  if (!all(sapply(dta[, ..tname], is.numeric))) {
    stop("tname = ",tname,  " is not numeric. Please convert it")
  }

  # Check if there is only 2 time periods
  tlist <- unique(dta[[tname]])[base::order(unique(dta[[tname]]))]
  if (length(tlist) != 2) {
    stop("The type of ddd specified only allow for two time periods (pre and post). Change type of ddd for multiple time periods")
  }

  # Flag for gname
  if ( !is.element(gname, base::colnames(dta))) {
    stop("gname = ",gname,  " could not be found in the data provided.")
  }

  # Check if there is only 2 groups
  if (dta[, uniqueN(get(gname))] != 2) {
    stop("The type of ddd specified only allow for two groups (treated and untreated). Change type of ddd for multiple groups")
  }



  # Flag for pname
  if ( !is.element(pname, base::colnames(dta))) {
    stop("pname = ",pname,  " could not be found in the data provided.")
  }


  # check if pname is numeric
  if (!all(sapply(dta[, ..pname], is.numeric))) {
    stop("pname = ",pname,  " is not numeric. Please convert it")
  }

  # Check if partition values are binary
  plist <- unique(dta[[pname]])[base::order(unique(dta[[pname]]))]
  if (length(plist) != 2) {
    stop("pname =", pname, " must have only two values (0 and 1). Please check pname")
  }

  # Check if idname is in the data
  if ( !is.element(idname, base::colnames(dta))) {
    stop("idname = ",idname,  " could not be found in the data provided.")
  }

  #  check if idname is numeric
  if (!all(sapply(dta[, ..idname], is.numeric))) {
    stop("data[, idname] must be numeric. Please convert it.")
  }

  # Check if any combination of idname and tname is duplicated
  n_id_year = anyDuplicated(dta[, .(idname, tname)])
  # If any combination is duplicated, stop execution and throw an error
  if (n_id_year > 0) {
    stop("The value of idname must be unique (by tname)")
  }

  # Check if partition is unique by idname
  checkPartitionUniqueness(dta, idname, pname)

  # Check if gname is unique by idname
  checkTreatmentUniqueness(dta, idname, gname)

  # Flag for weightsname
  if(!is.null(weightsname)){
    if ( !is.element(weightsname, base::colnames(dta))) {
      stop("weightsname = ",weightsname,  " could not be found in the data provided.")
    }
  }

  # FLAGS FOR DML ESTIMATION
  # if (est_method=="dml") {
  #   if (is.null(learners)) {
  #     stop("learners should be provided when est_method = 'dml'. Please, check documentation.")
  #   }
  #
  #   if (is.null(n_folds)) {
  #     stop("n_folds should be provided when est_method = 'dml'. Please, check documentation.")
  #   }
  #
  #   # check if there 2 learners provided in learners list
  #   if (length(learners) != 2) {
  #     stop("learners must be a list of two learners: ml_pa and ml_md. Please, check documentation.")
  #   }
  #
  #   # check if learner is a classifier type
  #   if (!inherits(learners$ml_pa, "LearnerClassif")) {
  #     stop("The learner must be a classification learner.")
  #   }
  #
  #   # check if learner  is a regression type
  #   if (!inherits(learners$ml_md, "LearnerRegr")) {
  #     stop("The learner must be a regression learner.")
  #   }
  # }

}


validate_args_multPeriods <- function(args, dta){

  # get args
  yname <- args$yname
  tname <- args$tname
  idname <- args$idname
  gname <- args$gname
  control_group <- args$control_group
  pname <- args$pname
  xformla <- args$xformla
  est_method <- args$est_method
  learners <- args$learners
  n_folds <- args$n_folds
  weightsname <- args$weightsname
  boot <- args$boot
  nboot <- args$nboot
  base_period <- args$base_period

  # Flag for based period: not in c("universal", "varying"), stop
  if (!base_period %in% c("universal", "varying")) {
    stop("base_period must be either 'universal' or 'varying'.")
  }

  # Flag for control group types
  if(!(control_group %in% c("nevertreated","notyettreated"))){
    stop("control_group must be either 'nevertreated' or 'notyettreated'")
  }

  # Flag for yname
  if (!is.element(yname, base::colnames(dta))) {
    stop("yname = ",yname,  " could not be found in the data provided.")
  }

  # Flag for tname
  if (!is.element(tname, base::colnames(dta))) {
    stop("tname = ",tname,  " could not be found in the data provided.")
  }

  # check if times periods are numeric
  if (!all(sapply(dta[, ..tname], is.numeric))) {
    stop("tname = ",tname,  " is not numeric. Please convert it")
  }

  # Flag for gname
  if ( !is.element(gname, base::colnames(dta))) {
    stop("gname = ",gname,  " could not be found in the data provided.")
  }

  # check if gname is numeric
  if (!all(sapply(dta[, ..gname], is.numeric))) {
    stop("gname = ",gname,  " is not numeric. Please convert it")
  }

  # Flag for pname
  if ( !is.element(pname, base::colnames(dta))) {
    stop("pname = ",pname,  " could not be found in the data provided.")
  }

  # check if pname is numeric
  if (!all(sapply(dta[, ..pname], is.numeric))) {
    stop("pname = ",pname,  " is not numeric. Please convert it")
  }

  # Check if partition values are binary
  plist <- unique(dta[[pname]])[base::order(unique(dta[[pname]]))]
  if (length(plist) != 2) {
    stop("pname =", pname, " must have only two values (0 and 1). Please check pname")
  }

  # Check if idname is in the data
  if ( !is.element(idname, base::colnames(dta))) {
    stop("idname = ",idname,  " could not be found in the data provided.")
  }

  #  check if idname is numeric
  if (!all(sapply(dta[, ..idname], is.numeric))) {
    stop("data[, idname] must be numeric. Please convert it.")
  }

  # Check if any combination of idname and tname is duplicated
  n_id_year = anyDuplicated(dta[, .(idname, tname)])
  # If any combination is duplicated, stop execution and throw an error
  if (n_id_year > 0) {
    stop("The value of idname must be unique (by tname)")
  }

  # Flag for weightsname
  if(!is.null(weightsname)){
    if ( !is.element(weightsname, base::colnames(dta))) {
      stop("weightsname = ", weightsname, " could not be found in the data provided.")
    }
  }

  # Faster and useful checks to make sere we have a well-balanced panel.

  # Check if partition is unique by idname
  checkPartitionUniqueness(dta, idname, pname)

  # Check if gname is unique by idname
  checkTreatmentUniqueness(dta, idname, gname)

  #TODO: ADD MORE CHECKS IF NEEDED FOR DML

}