library(magrittr)

selectedYear <- 2014

getWorldBankData <- function(indicatorCode, indicatorName) {
    baseName <- paste('API', indicatorCode, 'DS2_en_csv_v2', sep='_')
    ## Download zipfile if necessary
    zipfile <- paste(baseName, 'zip', sep='.')
    if (!file.exists(zipfile)) {
        zipurl <- paste(paste('http://api.worldbank.org/v2/en/indicator',
                              indicatorCode, sep='/'),
                        'downloadformat=csv', sep='?')
        download.file(zipurl, zipfile)
    }
    csvfile <- paste(baseName, 'csv', sep='.')
    ## This produces a warning because of the trailing commas. Safe to ignore.
    readr::read_csv(unz(zipfile, csvfile), skip=4,
                    col_types = list(`Indicator Name` = readr::col_character(),
                                     `Indicator Code` = readr::col_character(),
                                     `Country Name` = readr::col_character(),
                                     `Country Code` = readr::col_character(),
                                     .default = readr::col_double())) %>%
        dplyr::select(-c(`Indicator Name`, `Indicator Code`, `Country Name`))
}

## Obtain and merge World Bank data
worldBankData <-
    dplyr::left_join(
               getWorldBankData('MS.MIL.XPND.GD.ZS') %>%
               tidyr::gather(-`Country Code`, convert=TRUE,
                             key='Year', value=`Military expenditure (% of GDP)`,
                             na.rm = TRUE),
               getWorldBankData('NY.GDP.MKTP.CD') %>%
               tidyr::gather(-`Country Code`, convert=TRUE,
                             key='Year', value=`GDP (current US$)`,
                             na.rm = TRUE)) %>%
    dplyr::mutate(`Military expenditure (current $US)` =
                      `Military expenditure (% of GDP)`*`GDP (current US$)`/100) %>%
    dplyr::filter(Year == selectedYear) %>%
    dplyr::mutate(Year = NULL)

## Plotting: Obtain Geographic data
mapData <- tibble::as.tibble(ggplot2::map_data("world")) %>%
    dplyr::mutate(`Country Code` =
                      countrycode::countrycode(region, "country.name", "iso3c"),
                  ## This produces a warning but I do not see how we could do better
                  ## since we started with fuzzy names.
                  region = NULL, subregion = NULL)

combinedData <- dplyr::left_join(mapData, worldBankData)

## The default out-of-bounds function `censor` replaces values outside
## the range with NA. Since we have properly labelled the legend, we can
## project them onto the boundary instead
clamp <- function(x, range = c(0, 1)) {
    lower <- range[1]
    upper <- range[2]
    ifelse(x > lower, ifelse(x < upper, x, upper), lower)
}

ggplot2::ggplot(data = combinedData, ggplot2::aes(long,lat)) +
    ggplot2::geom_polygon(ggplot2::aes(group = group,
                                       fill  = `Military expenditure (current $US)`),
                          color = '#606060', lwd=0.05) +
    ggplot2::scale_fill_gradientn(colours= rev(viridis::magma(256, alpha = 0.5)),
                                  name   = "Million\nUS Dollars",
                                  trans  = "log",
                                  oob    = clamp,
                                  breaks = c(1e7,1e8,1e9,1e10,1e11,1e12),
                                  labels = c('\u2264 10', '100', '1K',
                                             '10K', '100K', '\u2265 1M'),
                                  limits = c(1e7,1e12)) +
    ggplot2::coord_fixed() +
    ggplot2::theme_bw() +
    ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5),
          axis.title = ggplot2::element_blank(),
          axis.text = ggplot2::element_blank(),
          axis.ticks = ggplot2::element_blank(),
          panel.grid.major = ggplot2::element_blank(),
          panel.grid.minor = ggplot2::element_blank(),
          panel.border = ggplot2::element_blank(),
          panel.background = ggplot2::element_blank()) +
    ggplot2::labs(title = paste("Countries by military expenditures in",
                                selectedYear))

ggplot2::ggsave(paste(selectedYear, 'militrary_expenditures_absolute.svg', sep='_'),
                height=100, units='mm')