na_count.R

# Load library   ------------------------------------------------------------
library(dplyr)

# Basic functions to explore NA in Dataframe  --------------------------------------------------------------


# Identify all missing variables per columns 
countNaCol <- function(data) {
  result <- vapply(data,function(x) sum(is.na(x)),integer(1))
  result <- data.frame(Nbna = result, napercentage = round(result/nrow(data),3))
  names(result) <- c('Nbna', 'napercentage')
  return(result)
}

# Identify missing Columns (only na values)
naColumns <- function(data) names(data)[vapply(data,
                            function(x) sum(is.na(x)) == nrow(data),
                            logical(1))]


# Identify many missing 
naColumnsM <- function(data,a=0.7) names(data)[vapply(data,
                                   function(x) sum(is.na(x)) > a * nrow(data),
                                   logical(1))]

# nb of na per row 
countNaRow <- function(data) rowSums(mutate_each(data,funs(is.na)))/ncol(data)


naRows <- function(data) which(countNaRow(data) == 1)
#all complete rows ( no na values )

completeRows <- function(data) which(countNaRow(data) == 0)