# Inequality Key Figures - replication with R and `lissyrtools`

library(lissyrtools)
library(tidyverse)


# 0) Load Data and Prepare it 
data <- lissyuse(
  data = "lu",   # Examples: `data = c("de", "it")`; `data = c("it20", "it16")`; Run: get_countries_lis() to check other iso2 codes, or get_years_lis(c("it", "es")) to check which years are available.
  vars = c("dhi", "nhhmem65", "nhhmem17","sex", "relation", "hpartner")
) %>%
  map(~ .x %>% 
        filter(!is.na(dhi)) %>%
        filter(relation == 1000) %>%
        mutate(
          new_wgt = hwgt*nhhmem, 
          new_chld_wgt = hwgt * nhhmem17,
          new_eldr_wgt = hwgt * nhhmem65 
        )
  ) %>%
  apply_iqr_top_bottom_coding("dhi", "hwgt", type = "type_4") %>% 
  apply_sqrt_equivalisation("dhi")


medians <- map(
  run_weighted_percentiles(data, "dhi", "new_wgt", type = "type_4"), 
  "50%"
)





# Relative Poverty Rates for the Total Population, Children and the Elderly (h-level data)


for (i in c(0.4, 0.5, 0.6)) {
  
  assign("poverty_line", map(medians  , ~ .x  * i)) # multiply the median by (0.4, 0.5, or 0.6), thus defining a poverty line
  
  data_chld_eldr <- map2(
    data,
    poverty_line,
    ~ .x %>%
      mutate(
        line = .y,  
        poor = if_else(dhi < line, 100, 0)
      )
  )
  
  assign(paste0("pvr_totpop_", i), # pvr_totpop_40, pvr_totpop_50, pvr_totpop_60
         run_weighted_mean(data_chld_eldr, "poor", "new_wgt"))
  
  assign(paste0("pvr_chld_", i), # pvr_chld_40, pvr_chld_50, pvr_chld_60
         run_weighted_mean(data_chld_eldr, "poor", "new_chld_wgt"))
  
  assign(paste0("pvr_eldr_", i), # pvr_eldr_40, pvr_eldr_50, pvr_eldr_60
         run_weighted_mean(data_chld_eldr, "poor", "new_eldr_wgt"))
}






# Collect them into a list
all_stats <- list(
  poorAll4 = pvr_totpop_0.4,
  poorAll5 = pvr_totpop_0.5,
  poorAll6 = pvr_totpop_0.6,
  poorK4 = pvr_chld_0.4,
  poorK5 = pvr_chld_0.5,
  poorK6 = pvr_chld_0.6,
  poorE4 = pvr_eldr_0.4,
  poorE5 = pvr_eldr_0.5,
  poorE6 = pvr_eldr_0.6
)

results <- map(all_stats,  ~ structure_to_plot(.x, print_columns = FALSE)) 

final <- imap(results , ~ .x %>% mutate(short_ikf = .y)) %>% 
  bind_rows() %>% 
  arrange(short_ikf, dname) %>%  
  mutate(
    indicator = case_when(
      short_ikf == "poorAll4"   ~ "Relative Poverty Rates - Total Population (40%)",
      short_ikf == "poorAll5"   ~ "Relative Poverty Rates - Total Population (50%)",
      short_ikf == "poorAll6"   ~ "Relative Poverty Rates - Total Population (60%)",
      short_ikf == "poorK4"     ~ "Relative Poverty Rates - Children (40%)",
      short_ikf == "poorK5"     ~ "Relative Poverty Rates - Children (50%)",
      short_ikf == "poorK6"     ~ "Relative Poverty Rates - Children (60%)",
      short_ikf == "poorE4"     ~ "Relative Poverty Rates - Elderly (40%)",
      short_ikf == "poorE5"     ~ "Relative Poverty Rates - Elderly (50%)",
      short_ikf == "poorE6"     ~ "Relative Poverty Rates - Elderly (60%)"
    )
  ) %>%
  rename(value_ikf_R = value)

write.csv(final)