data-wrangling.qmd

---
title: "Data Wrangling"
format: html
---

Run the code chunks in this file to wrangle the data for the app. 

## Chunk 1

```{r}
# Download the selected indicators for the post-1990 period.
library(vdemdata)
library(dplyr)
library(readr)
library(countrycode)

vdem_df <- vdem |>
  filter(year >= 1990) |>
  select(
    country_name,
    country_id,
    year, 
    v2x_polyarchy,
    v2x_libdem,
    v2x_partipdem,
    v2x_delibdem,
    v2x_egaldem,
    v2x_civlib,
    v2x_rule,
    v2xnp_client,
    v2xnp_regcorr,
    v2x_gender,
    e_gdppc,
    region = e_regionpol_6C) |>
    mutate(gdp_pc = scales::dollar_format()(e_gdppc*1000),
           e_gdppc = e_gdppc*1000) |>
    mutate(
    region = case_match(region,
                     1 ~ "Eastern Europe", 
                     2 ~ "Latin America",  
                     3 ~ "Middle East",   
                     4 ~ "Africa", 
                     5 ~ "The West", 
                     6 ~ "Asia")
  ) |>
  # add iso3c code
  mutate(iso3c = countrycode(sourcevar = country_id, 
        origin = "vdem",         
        destination = "wb"))  |> 
  relocate(iso3c, .after = country_id) 

glimpse(vdem_df)

#write_rds(vdem_tbl, "vdem_trends.rds")
```

## Chunk 2 

```{r}
# Now merge these data with country shapes from `rnaturalearth`.
library(rnaturalearth)
library(sf)
library(dplyr)
library(tidyr)

# Load world boundaries
world <- ne_countries(scale = "medium", returnclass = "sf") |> 
  select(iso_a3, geometry) |> 
  filter(iso_a3 != "ATA") |>
  rename(iso3c = iso_a3)

# Expand vdem_df to include all country-year combinations
years <- seq(min(vdem_df$year), max(vdem_df$year))
countries <- unique(world$iso3c)

expanded_vdem_df <- expand.grid(iso3c = countries, year = years) |> 
  left_join(vdem_df, by = c("iso3c", "year"))

# Join with world boundaries and abbreviate country names
vdem_data <- left_join(world, expanded_vdem_df, by = "iso3c") |>
  mutate(country_name = if_else(iso3c == "GRL", "Greenland", country_name)) |>
  mutate(country_name = if_else(iso3c == "SSD", "South Sudan", country_name)) |>
  mutate(country_name = if_else(country_name == "North Macedonia", "N. Macedonia", country_name)) |> 
  mutate(country_name = if_else(country_name == "Bosnia and Herzegovina", "BiH", country_name)) |>
  mutate(country_name = if_else(country_name == "Dominican Republic", "Dominican Rep.", country_name)) |>
  mutate(country_name = if_else(country_name == "Trinidad and Tobago", "T&T", country_name)) |>
  mutate(country_name = if_else(country_name == "United Arab Emirates", "UAE", country_name)) |>
  mutate(country_name = if_else(country_name == "Palestine/Gaza", "Palestine", country_name)) |>
  mutate(country_name = if_else(country_name == "Equitorial Guinea", "Eq. Guinea", country_name)) |>
  mutate(country_name = if_else(country_name == "Republic of the Congo", "Congo-Brazzaville", country_name)) |>
  mutate(country_name = if_else(country_name == "Democratic Republic of the Congo", "DRC", country_name)) |>
  mutate(country_name = if_else(country_name == "Central African Republic", "CAR", country_name)) |>
  mutate(country_name = if_else(country_name == "Sao Tome and Principe", "STP", country_name)) |>
  mutate(country_name = if_else(country_name == "United States of America", "United States", country_name)) |>
  mutate(country_name = if_else(country_name == "Burma/Myanmar", "Myanmar", country_name)) |>
  mutate(country_name = if_else(country_name == "Papua New Guinea", "PNG", country_name)) 
  
# Replace missing e_gdppc and gdp_pc values with previous year's values
# Note: could try solving this issue by mergin WB data instead
vdem_data <- vdem_data |> 
  group_by(iso3c) |> 
  mutate(
    e_gdppc = if (any(!is.na(e_gdppc))) zoo::na.locf(e_gdppc, na.rm = FALSE) else e_gdppc,
    gdp_pc = if (any(!is.na(gdp_pc))) zoo::na.locf(gdp_pc, na.rm = FALSE) else gdp_pc
  ) |> 
  ungroup()


glimpse(vdem_data)

write_rds(vdem_data, "vdem_data.rds")
```