naniar package

The missing data can be checked using functions included in the naniar package (see also packages VIM and mice). There are functions that are useful for structural understanding and summarization of missing values. The following is a list of functions implemented in this package:

• n_var_miss() returns number of missing variables.
• miss_var_summary() creates a summary table of missing variables, their proportions, for each variable.
• miss_case_summary() creates a summary table of the number of misses and the proportion of misses for each individual.
• miss_var_span() creates a summary table of the number of missings in a given span, or breaksize, for a single selected variable.
• geom_miss_point() visualizes missing values on the same axes as the regular values, colours the missing and not missing points.
• gg_miss_var() visualizes miss_var_summary() function results.
• gg_miss_case() visualizes miss_case_summary() function results.
• gg_miss_span() visualizes miss_var_span() function results.
• gg_miss_case_cumsum() visualizes cumulative sum of missing values, reading the rows of the dataset from the top to bottom.
• gg_miss_var_cumsum() visualizes cumulative sum of missing values, reading columns from the left to the right of your dataframe.
• gg_miss_upset() visualizes combinations of missingness and intersections of missingness amongst variables.
• gg_miss_fct() visualizes number of missings in each column, broken down by a categorical variable from the dataset.
• gg_miss_which() visualizes a set of rectangles that indicate whether there is a missing element in a column or not.
• vis_miss() visualizes missing status with individuals/variables.

Below is an example of applying these functions to airquality data:

# Libraries
library(naniar)
library(ggplot2)

# Create data
set.seed(123)
data(airquality)
airquality$Temp[sample(1:nrow(airquality), size=10)] <- NA
airquality$Day[sample(1:nrow(airquality), size=50)] <- NA

# Functions
#n_var_miss(airquality)
miss_var_summary(airquality)

miss_case_summary(airquality)

miss_var_span(airquality, Day, span_every = 25)

geom_miss_point() gg_miss_var() gg_miss_case() gg_miss_span() gg_miss_case_cumsum() gg_miss_var_cumsum() gg_miss_upset() gg_miss_fct() gg_miss_which() vis_miss()

vis_miss1 function

Last update date : 08/08/2024

Download vis_miss1.R

This function is an update of vis_miss function that visualizes a dataframe to display missingness. Additional arguments have been added, for example to flip the coordinates (observations in columns rather than rows). The arguments of this function are given below (additional arguments and those of vis_miss). It returns a ggplot of the missingness inside a dataframe, colouring cells according to missingness, where black indicates a missing cell and grey indicates a present cell. As it returns a ggplot object, it is very easy to customize and change labels.

vis_miss1(
x, a dataframe.
cluster = FALSE, logical. TRUE specifies that you want to use hierarchical clustering (mcquitty method) to arrange rows according to missingness. FALSE specifies that you want to leave it as is. Default value is FALSE.
sort_miss = FALSE, logical. TRUE arranges the columns in order of missingness. Default value is FALSE.
show_perc = TRUE, logical. TRUE now adds in the % of missing/complete data in the whole dataset into the legend. Default value is TRUE.
show_perc_col = TRUE, logical. TRUE adds in the % missing data in a given column into the x axis. Can be disabled with FALSE. Default value is TRUE.
large_data_size = 9e+05, integer default is 900000 (given by nrow(data.frame) * ncol(data.frame)).
warn_large_data = TRUE, logical - warn if there is large data.
flip = FALSE, flip the coordinates if TRUE (observations in columns).
vect.var = NULL, variable names.
vect.names = NULL, variables names we want to display.
varassoc = NULL, “parent” variables that imply nested responses. Put “” if no “parent” variable.
codassoc = NULL, code that would lead to the response of a nested variable. obs = NULL variable name for observations.
)

Here is an example where optional parameters are used:

# Libraries
library(naniar)


# Create data
set.seed(8)
data <- data.frame(
  ID = as.character(1:150)
)

len <- sample(1:50, size=50, replace=TRUE)
for(i in 1:50){
  ind <- sample(1:nrow(data), size=len[i])
  data[,paste0("VAR_",i)] <- sample(1:50, size=nrow(data), replace=TRUE)
  data[,paste0("VAR_",i)][ind] <- NA
}


# Plot
vect_var <- names(data)[-1]
vect_nom <- names(data)[-1]
# VAR_2 if VAR_1=12
var_assoc <- rep("",length(vect_var))
var_assoc[vect_var%in%"VAR_2"] <- "VAR_1"
cod_assoc <- rep("",length(vect_var))
cod_assoc[vect_var%in%"VAR_2"] <- "12"
plot_miss <- vis_miss1(data, sort_miss = FALSE, flip=TRUE, #obs="ID",
                       vect.var = vect_var, vect.names = vect_nom,
                       varassoc = var_assoc, codassoc = cod_assoc)