inst/helpers.R

In Kafi1993/Explore: Explore

# Hilfs-FUnktionen

extractNumeric <- function(dframe){
  dframe[sapply(dframe, is.numeric)] 
}



naCount <- function(dframe){
  sapply(dframe, function(column){sum(is.na(column))})
}



meanAll <- function(dframe){
  round(sapply(dframe, mean, na.rm = T), digits = 2)
}



medianAll <- function(dframe){
  round(sapply(dframe, median, na.rm = T), digits = 2)
} 



sdAll <- function(dframe){
  round(sapply(dframe, sd, na.rm = T), digits = 2)
}


minAll <- function(dframe){
  round(sapply(dframe, min, na.rm = T), digits = 2)
}


maxAll <- function(dframe){
  round(sapply(dframe, max, na.rm = T), digits = 2)
}


skewAll <- function(dframe){
  round(sapply(dframe, skew, na.rm = T), digits = 2)
}


kurtosisAll <- function(dframe){
  round(sapply(dframe, kurtosi, na.rm = T), digits = 2)
}

# Hauptfuntktion
getDescriptives <- function(rawdata){
  numericData <- extractNumeric(rawdata)
  
  li <- list()
  
  li$na <- naCount(numericData)
  li$mean <- meanAll(numericData)
  li$median <- medianAll(numericData)
  li$sd <- sdAll(numericData)
  li$min <- minAll(numericData)
  li$max <- maxAll(numericData)
  li$kurtosis <- kurtosisAll(numericData)
  li$skew <- skewAll(numericData)
  
  return(li)
}


###changes structure of the dataframe to make it usable for ggplot

convertData <- function(rawdata){
  stacked_data <- stack(rawdata)
  # converts columns to factors, so that it can be used by ggplot as a categorical variable
  as.data.frame(sapply(stacked_data, as.factor))
}


divideByPattern <- function(dframe, pattern, n.col = 15){
  if(is.character(pattern) == TRUE){
    dframe <- dframe[,grep(pattern, names(dframe), value=T)]
  }
  else if(is.numeric(pattern) == TRUE){
    dframe <- dframe[,pattern]
  }
  else{
    stop('"pattern" must only contain character values or vectorial index-values.')
  }
  
  dframe <- extractNumeric(dframe)
  if(NCOL(dframe)>n.col){
    dframe <- dframe[,1:n.col]
  }
  return(dframe)
}


# Plot-function

stackedBars <- function(dframe, type = position_fill, pattern = "", n.col = 15, palette = "Blues", direction = 1){
  
  dframe <- divideByPattern(dframe = dframe, pattern = pattern, n.col = n.col)
  dframe <- convertData(dframe)
  
  # ind codes for the items - shown on x-axis
  # values contains the responses of participants/values of the items
  # see function "convertData" for details
  ggplot(data = dframe, mapping = aes(x = ind, fill = values)) +
    geom_bar(position = type(reverse = TRUE)) +
    coord_flip() +
    labs(x = " ") +
    scale_fill_brewer(palette = palette, direction = direction)
}


parallelBars <- function(dframe, position = position_fill(), pattern = "", n.col = 15, palette = "Blues", direction = 1){
  
  dframe <- divideByPattern(dframe = dframe, pattern = pattern, n.col = n.col)
  dframe <- convertData(dframe)
  
  # ind codes for the items - shown on x-axis
  # values contains the responses of participants/values of the items
  ggplot(data = dframe, mapping = aes(x = ind, fill = values)) +
    geom_bar(position = position) +
    labs(x = " ") +
    scale_fill_brewer(palette = palette, direction = direction) +
    guides(fill = guide_legend(reverse = T))
}