Change gg_miss_case to be more practical

[njtierney]

The idea behind gg_miss_case is to give you a sense of how many missings there are in your dataset when we look across the cases (the rows).

However, I'm fairly convinced that the current plotting method (with points and lines) isn't very nice, so I decided to change the plotting method to remove the unnecessary points, and it also comes with a reordering method, where it gives you a sense of how many missings there across cases, ordering from most to least missing.

# ggobject <- 
library(tidyverse)
library(naniar)
gg_miss_case2 <- function(data, order_rows = FALSE){
  
  if (order_rows) {
    miss_case_summary(airquality, order = TRUE) %>%
      # overwrite case
      mutate(case = 1:n()) %>%
      ggplot(data = .,
             aes(x = case,
                 y = n_miss)) +
      geom_col(width = 1,
               colour = "grey",
               fill = "grey") + 
      coord_flip() + 
      labs(y = "# Missing",
           x = "# Cases") +
      theme_minimal() + 
      scale_x_reverse()
    
  } else if (!order_rows) {
  ggplot(data = miss_case_summary(data),
         aes(x = case,
             y = n_miss)) +
  geom_col(width = 1,
           colour = "grey",
           fill = "grey") +
  coord_flip() +
  labs(y = "# Missing",
       x = "Cases") +
  theme_minimal() + 
  scale_x_reverse()
  }
}
# return(ggobject)

gg_miss_case(airquality)

gg_miss_case2(airquality)

gg_miss_case2(airquality, order_rows = TRUE)

[njtierney]

Current output, from branch - gg-miss-groups

library(naniar)

gg_miss_case(airquality)

gg_miss_case(airquality, order_cases = TRUE)

gg_miss_case(airquality, group = Month)