Datasets

repo = "https://github.com/fivethirtyeight/data/tree/master/terrorism"
article = "https://fivethirtyeight.com/features/the-rise-of-religiously-inspired-terrorism-in-france"
link = "https://raw.githubusercontent.com/fivethirtyeight/data/master/terrorism/eu_terrorism_fatalities_by_country.csv"

terror = read.csv(link) %>%
  dplyr::rename(Year = iyear)

Overview

By some definitions, an act of violence becomes labeled as terrorism when it is meant to inflict fear. The 1970’s saw a significant rise in terrorism across several countries in the EU. The most significant death tolls between 1970 to 2014 were observed in the UK and Spain.

LogTerror <- function(df, log_base = exp(1)) {
  df %>%
    log(., log_base) %>%
      { .[df == 0] = 0 ; . } %>%
        { .[, 1] = df[, 1] ; .}
}

MeltTerror <- function(df, reference = terror, x_var = c("Year")) {
  sapply(reference, sum) %>%
    { which(. < 100) } %>%
      { . * -1 } %>%
        df[, .] %>%
          reshape2::melt(., id = x_var, na.rm = TRUE)
}

RenameTerror <- function(df) {
  dplyr::rename(df, Country = variable, Fatalities = value)
}

Exploratory Analysis

Shown below is a representation of the fatalities per year for six countries that exceeded 100 deaths.

terror %>%
  LogTerror(.) %>%
    MeltTerror(.) %>%
      RenameTerror(.) %>%
        ggplot2::ggplot(.) +
        ggplot2::aes(x = Year, y = Fatalities, fill = Country) +
        ggplot2::geom_area() +
        ggplot2::theme_classic()

Findings and Recommendations

It is curious that these two large countries, Italy and Ireland, have not had any cases since 1995. It would be worthwhile to see if there are social, political, or economic causes for this remedial action.

terror %>%
  LogTerror(.) %>%
    MeltTerror(.) %>%
      RenameTerror(.) %>%
        .[.$Country %in% c("Italy", "Ireland"), ] %>%
          ggplot2::ggplot(.) +
          ggplot2::aes(x = Year, y = Fatalities, group = Country, color = Country) +
          ggplot2::geom_line(size = 1.1) +
          ggplot2::theme_classic()