Plot Missingness for a Date Variable

Plot Missingness for a Date Variable

plot_date_variable_missingness(df = NULL, var = NULL,
  start_year = NULL, end_year = NULL, split_year = NULL,
  facet_by_year = TRUE)

Arguments

df	A dataframe containing cleaned ETS data as produced by tbinenglanddataclean.
var	A character string indicating the name of the date variable to explore missingness within.
start_year	Numeric, indicating which year to start including data from (inclusive)
end_year	Numeric, indicating which year to stop including data from (exclusive)
split_year	The year to use as a splitting point when facetting. By default uses the mean year.
facet_by_year	Logical defaults to TRUE. Should the plots be facetted by a year midpoint

Value

A list of plots including: missing data by month and by day.

Examples

## Code
plot_date_variable_missingness
#> function (df = NULL, var = NULL, start_year = NULL, end_year = NULL, 
#>     split_year = NULL, facet_by_year = TRUE) 
#> {
#>     year_facet <- NULL
#>     notifications <- NULL
#>     year_strat <- NULL
#>     nn <- NULL
#>     df$date <- df[[var]]
#>     if (is.null(start_year)) {
#>         start_year <- df$date %>% year %>% min(na.rm = TRUE)
#>     }
#>     if (is.null(end_year)) {
#>         end_year <- df$date %>% year %>% max(na.rm = TRUE)
#>     }
#>     df_count <- df %>% filter(year(date) >= start_year, year(date) <= 
#>         end_year) %>% drop_na(date) %>% count(date, .drop = FALSE, 
#>         name = "notifications") %>% mutate(notifications = notifications %>% 
#>         replace_na(0))
#>     years_of_data <- df_count$date %>% year() %>% unique() %>% 
#>         as.numeric()
#>     if (is.null(split_year)) {
#>         split_year <- years_of_data %>% mean(na.rm = TRUE) %>% 
#>             floor
#>     }
#>     df_count <- df_count %>% dplyr::mutate(year_strat = cut(year(date) %>% 
#>         as.integer, breaks = c(min(years_of_data) - 1, split_year, 
#>         max(years_of_data) + 1), labels = c(paste0(min(years_of_data), 
#>         "-", split_year - 1), paste0(split_year, "-", max(years_of_data))), 
#>         right = FALSE))
#>     if (facet_by_year) {
#>         df_count <- df_count %>% group_by(year_strat)
#>     }
#>     month_plot <- df_count %>% mutate(date = floor_date(date, 
#>         "month")) %>% count(date, wt = notifications) %>% add_count(year(date), 
#>         wt = n, name = "nn") %>% mutate(n = n/nn) %>% mutate(month = month(date, 
#>         label = TRUE)) %>% ggplot(aes(x = month, y = n)) + geom_violin(draw_quantiles = c(0.25, 
#>         0.5, 0.75)) + geom_jitter(alpha = 0.2) + scale_y_sqrt(labels = percent) + 
#>         theme_minimal() + labs(x = "Month", y = "Percentage of annual notifications (sqrt)")
#>     day_plot <- df_count %>% mutate(date = floor_date(date, "day")) %>% 
#>         count(date, wt = notifications) %>% add_count(floor_date(date, 
#>         "month"), wt = n, name = "nn") %>% mutate(n = n/nn) %>% 
#>         mutate(mday = mday(date)) %>% ggplot(aes(x = mday, y = n, 
#>         group = mday)) + geom_violin(draw_quantiles = c(0.25, 
#>         0.5, 0.75)) + geom_jitter(alpha = 0.05) + scale_y_sqrt(labels = percent) + 
#>         scale_x_continuous(minor_breaks = NULL, breaks = seq(1, 
#>             31, 2)) + theme_minimal() + labs(x = "Day of the month", 
#>         y = "Percentage of monthly notifications (sqrt)")
#>     plots <- list(month_plot, day_plot)
#>     if (facet_by_year) {
#>         plots <- plots %>% map(~. + facet_wrap(~year_strat, scales = "free_y"))
#>     }
#>     names(plots) <- c("by_month", "by_day")
#>     return(plots)
#> }
#> <bytecode: 0x55630a8840a0>
#> <environment: namespace:ETSMissing>