R/crit_quant.R

In bibliorefer: Generator of Main Scientific References

#' Function of quantitative validation criteria
#'
#' This function quantitatively validates the list of main scientific references obtained. The function input is a list object with the two dataframes generated by the principal_lister function. The crit_quant function returns a list of two dataframes with the quantitative criteria that validate the list of main scientific references obtained.
#'
#' @param input_ent input_ent is a dataframe with the list of main scientifics references generated by the principal_lister function
#'
#' @importFrom "stats" "na.omit"
#'
#' @return The function return a list object with two dataframes with quantitative criteria that validate the list of main scientifics references
#' @export
#'
#' @references
#'
#'1 - Asymptotic Theory of Statistics and Probability. Anirban DasGupta. New York: Springer Science+Business Media, LLC, 2008.
#'
#'2 - Quantitative science studies should be framed with middle-range theories and concepts from the social sciences Heinze, T.,  Jappe, A. Quantitative Science Studies, 1(3), pp. 983–992, 2020
#'
#' @examples
#'
#' #Call the principal_lister function
#' input_tam <- 100
#' file_db <- system.file("extdata","example_database.csv", package = "bibliorefer")
#' separator <- ","
#' input_date <- example_database(file_db, separator)
#' principal_refer <- principal_lister(input_date,input_tam)
#'
#' #Call the crit_quant function
#' quant_validation <- crit_quant(principal_refer)
#' quant_validation[[1]]
#' quant_validation[[2]]
#'

crit_quant <- function(input_ent){

  # Data entry parameters
  tabel_ent <- input_ent
  nume_classes <- 10

  # Function of quantitative criteria of the Author
  tabefai_autor <- function(tabel_ent,nume_classes){

    # Data entry parameters
    tabel_entautor <- tabel_ent[[1]]
    mposautor <- tabel_ent[[2]]$mposautor

    #Calculate the length of class extremes
    extrem <- nume_classes+1
    produt_autor <- tabel_entautor$`Productive position of author`
    produt_autor <- as.numeric(produt_autor)
    produt_autor <- cbind(produt_autor)

    # Calculate class length
    maxim_autor <- mposautor
    minim_autor <- min(produt_autor)
    tam_autor <- (maxim_autor-minim_autor)/nume_classes

    # Create frequency distribution classes
    autor_classes <- seq(minim_autor, maxim_autor,
                         length.out = extrem)
    autor_classes <- round(autor_classes)
    autor_classes <- cbind(autor_classes)

    # Create production range variables
    tama <- 10
    parte1 <- numeric(tama)
    parte2 <- numeric(tama)
    cami_entautor <- numeric(tama)

    # Start variables
    parte1[1] <- autor_classes[1]
    parte2[2] <- autor_classes[2]
    cami_entautor[1] <- paste(parte1[1], parte2[2],
                              sep = " - ")
    # Create the loop
    for(i in 3:tama){

      parte1[i-1] <- autor_classes[i-1] + 1
      parte2[i] <- autor_classes[i]
      cami_entautor[i-1] <- paste(parte1[i-1], parte2[i],
                                  sep = " - ")
    }
    parte1[tama] <- autor_classes[tama]+1
    parte2[tama+1] <- autor_classes[tama+1]
    cami_entautor[tama] <- paste(parte1[tama], parte2[tama+1],
                                 sep = " - ")

    # Create the matrix with productivity ranges
    cami_entautor <- cbind(cami_entautor)

    # Creates the frequency distribution of values
    fabs_autor <- table(cut(produt_autor,
                            include.lowest = T, breaks=autor_classes,
                            right=F, labels = cami_entautor))
    fabs_autor <- as.data.frame(fabs_autor)
    fabsac_autor <- cumsum(fabs_autor$Freq)

    #Calculate relative frequency
    tota_autor <- sum(fabs_autor$Freq)
    fr_autor <- 100*fabs_autor$Freq/tota_autor
    fr_autor <- round(fr_autor,digits = 3)

    #Calculates the accumulated frequency
    fac_autor <- cumsum(fr_autor)

    #Calculates the table with frequency distribution
    faix_autor <- c("One","Two","Three","Four","Five",
                    "Six","Seven","Eight","Nine","Ten")
    difre_autor <- cbind(faix_autor,cami_entautor,fabs_autor$Freq,
                         fabsac_autor,fr_autor,fac_autor)
    difre_autor <- as.data.frame(difre_autor)
    # Column names
    colnames(difre_autor) <- c("Range", "Productive position",
                               "Number of authors",
                               "Accumulated number of authors",
                               "Percentage",
                               "Accumulated percentage")
    # Return results
    return(difre_autor)

  }

  # Function of the quantitative criteria of the magazine
  tabefai_revist <- function(tabel_ent,nume_classes){

    # Data entry parameters
    tabel_entrevist <- tabel_ent[[1]]
    mposrevist <- tabel_ent[[2]]$mposrevist

    #Calculate the length of class extremes
    extrem <- nume_classes+1
    produt_revist <- tabel_entrevist$`Productive position of journal`
    produt_revist <- as.numeric(produt_revist)
    tabee_revist <- na.omit(produt_revist)

    # Calculate class length
    maxim_revist <- max(mposrevist)
    minim_revist <- min(tabee_revist)
    tam_revist <- (maxim_revist-minim_revist)/nume_classes

    # Create frequency distribution classes
    revist_classes <- seq(minim_revist, maxim_revist,
                          length.out = extrem)
    revist_classes <- round(revist_classes)
    revist_classes <- cbind(revist_classes)

    # Create production range variables
    tama <- 10
    parte1 <- numeric(tama)
    parte2 <- numeric(tama)
    cami_erevist <- numeric(tama)

    # Creates the frequency distribution of values
    parte1[1] <- revist_classes[1]
    parte2[2] <- revist_classes[2]
    cami_erevist[1] <- paste(parte1[1], parte2[2],
                             sep = " - ")
    # Create the loop
    for(i in 3:tama){

      parte1[i-1] <- revist_classes[i-1] + 1
      parte2[i] <- revist_classes[i]
      cami_erevist[i-1] <- paste(parte1[i-1], parte2[i],
                                 sep = " - ")
    }
    parte1[tama] <- revist_classes [tama]+1
    parte2[tama+1] <- revist_classes[tama+1]
    cami_erevist[tama] <- paste(parte1[tama], parte2[tama+1],
                                sep = " - ")

    # Create the matrix with productivity ranges
    cami_erevist <- cbind(cami_erevist)

    # Creates the frequency distribution of values
    fabs_revist <- table(cut(produt_revist,
                             include.lowest = T,
                             breaks=revist_classes,
                             right=F, labels = cami_erevist))
    fabs_revist <- as.data.frame(fabs_revist)
    fabsac_revist <- cumsum(fabs_revist$Freq)

    #Calculate relative frequency
    tota_revist <- sum(fabs_revist$Freq)
    fr_revist <- 100*fabs_revist$Freq/tota_revist
    fr_revist <- round(fr_revist,digits = 2)

    #Calculates the accumulated frequency
    fac_revist <- cumsum(fr_revist)

    #Calculates the table with frequency distribution
    faix_revist <- c("One","Two","Three","Four","Five",
                     "Six","Seven","Eight","Nine","Ten")
    difre_revist <- cbind(faix_revist,cami_erevist,
                          fabs_revist$Freq,
                          fabsac_revist,fr_revist,
                          fac_revist)
    difre_revist <- as.data.frame(difre_revist)
    # Column names
    colnames(difre_revist) <- c("Range", "Productive position",
                                "Number of magazines",
                                "Accumulated number of magazines",
                                "Percentage",
                                "Accumulated percentage")
    # Return results
    return(difre_revist)
  }

  # Call functions
  difreq_autor <- tabefai_autor(tabel_ent,nume_classes)
  difreq_revistr <- tabefai_revist(tabel_ent,nume_classes)

  # Create the list of results
  difreq_results <- list()
  difreq_results <- list(difreq_autor,difreq_revistr)
  names(difreq_results) <- c("Critquant_author",
                             "Critquant_journal")

  # Return results
  return(difreq_results)
}