R/get_type.R

In shinyr: Data Insights Through Inbuilt R Shiny App

#' getType
#' @title Get Type
#' @param vec A vector of any choice, to detect between numeric or character
#' @author Jayachandra N
#' @return type of the given vector
#' @export
#' @examples
#' getType(iris$Species)
#' getType(as.factor(c(1,0,1,1,0,NA,1, NULL)))
#' getType(as.factor(c(1, NULL,0,1,1,0,1,'a')))
#' getType(c(1,2,3,4, NA))
#' getType(letters[1:4])
getType <- function(vec) {
  cls <- class(vec)
  if(cls == "character") {
    return('character')
  } 
  
  if(cls == 'numeric' | cls == 'integer') {
    return('numeric')
  } 
  
  if(cls == 'factor') {
    tryCatch(
      {
        y <- y[!is.null(y)]
        y <- y[!is.na(y)]
        y <- y[y!='']
        y <- y[y!= ' ']
        y <- y[y!="NULL"]
        y <- y[y!="NA"]
        
        x <- sum(is.na(as.numeric(levels(y))))
        
        if(x == 0) {
          return('numeric')
        } else {
          return('character')
        }
      },
      
      error=function(cond) {
        return('character')
      })
  }
}

#' getTypeOfColumns
#' @author Jayachandra N
#' @param df data frame
#' @return Data frame of column name and it's type
#' @export
#' @examples
#' getTypeOfColumns(mtcars)
#' getTypeOfColumns(iris)
getTypeOfColumns <- function(df) {
  # df <- iris
  res <- lapply(names(df), FUN = function(x) {
    getType(df[[x]])
  })
  names(res) <- names(df)
  res <- as.data.frame(do.call('rbind', res))
  res$column <- names(df)
  res <- res[,c(2,1)]
  row.names(res) <- NULL
  names(res) <- c('Column', "Type")
  return(res)
}