War is over (if you want it?)

Two Points Make a Line

4 months ago

[This article was first published on Two Points Make a Line, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)

Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

People’s feelings about the Christmas season span a wide range, from very positive to very negative. As for where I fall, well:

< details> < summary>Code

library(tidyverse)
library(ggplot2)
library(lubridate)

df <- here::here("datasets", "ucdp-prio-acd-221.csv") %>%
  read_csv() %>%
  select(conflict_id, location, year, type_of_conflict, region, ep_end_date, region) %>%
  mutate(
    type_of_conflict = factor(type_of_conflict),
    xmas = as.Date(paste0(year, "-12-25")),
    count = ifelse(!(ep_end_date > xmas), 0, 1)
  ) %>%
  replace_na(list(count = 1))

conflicts <- df %>%
  group_by(year, type_of_conflict) %>%
  summarize(count = sum(count)) %>%
  ungroup()

ggplot(conflicts, aes(x = year, y = count, fill = type_of_conflict)) +
  geom_bar(stat = "identity", position = "stack", width = .7) +
  geom_hline(yintercept = 0, linewidth = .25, color = "black") +
  labs(
    title = "Another year over",
    subtitle = "Active armed conflicts as of Christmas Day, 1946-2021",
    caption = "Source: UCDP/PRIO Armed Conflict Dataset v.22.1"
  ) +
  scale_fill_manual(
    name = "Type of conflict",
    labels = c("Extrasystemtic", "Interstate", "Intrastate", "Internationalized intrastate"),
    values = rev(c("#1046b1", "#c52e9b", "#ff505b", "#ffa600"))
  ) +
  theme_minimal(base_family = "karla") +
  theme(
    plot.title = element_text(size = 16, face = "bold", hjust = .5),
    plot.subtitle = element_text(size = 14, hjust = .5, margin = margin(b = 15)),
    plot.caption = element_text(size = 8, hjust = 0, margin = margin(t = 10)),
    axis.ticks = element_blank(),
    axis.title = element_blank(),
    axis.text.x = element_text(size = 12, margin = margin(t = 5)),
    axis.text.y = element_text(size = 12, margin = margin(r = 5)),
    legend.position = "bottom",
    legend.title = element_text(size = 11, face = "bold", margin = margin(r = 10)),
    legend.text = element_text(size = 11),
    legend.key.size = unit(.4, "lines"),
    panel.background = element_rect(fill = "gray97", color = NA),
    panel.grid.major.x = element_blank(),
    panel.grid.major.y = element_line(color = "white"),
    panel.grid.minor.x = element_blank(),
    panel.grid.minor.y = element_blank()
  )

The data is from the UCDP/PRIO Armed Conflict Dataset. I’ll go through how I constructed the above chart.

First, some words about the dataset. Each observation is a conflict-year. A conflict is dated from its first battle-related death, but only active “episodes” — defined as 25 battle-related deaths — are recorded in the database. For example, the Basque conflict has the following entries:

here::here("datasets", "ucdp-prio-acd-221.csv") %>%
  read_csv() %>%
  select(conflict_id, side_a, side_b, year, start_date, start_date2, ep_end_date) %>%
  filter(conflict_id == "342") %>%
  knitr::kable(align = "c")

conflict_id	side_a	side_b	year	start_date	start_date2	ep_end_date
342	Government of Spain	ETA	1978	1968-06-07	1978-10-22	NA
342	Government of Spain	ETA	1979	1968-06-07	1978-10-22	NA
342	Government of Spain	ETA	1980	1968-06-07	1978-10-22	NA
342	Government of Spain	ETA	1981	1968-06-07	1978-10-22	NA
342	Government of Spain	ETA	1982	1968-06-07	1978-10-22	1982-12-29
342	Government of Spain	ETA	1985	1968-06-07	1985-12-23	NA
342	Government of Spain	ETA	1986	1968-06-07	1985-12-23	NA
342	Government of Spain	ETA	1987	1968-06-07	1985-12-23	1987-12-31
342	Government of Spain	ETA	1991	1968-06-07	1991-06-28	1991-12-13

While it is recorded as starting on 7 June 1968, its first active episode started on 22 October 1978 and lasted until 29 December 1982. A second episode occurred from 23 December 1985 to 31 December 1987, then a third from 28 June to 13 December 1991. For my purposes, this would count as an active conflict for the Christmas Days of 1978-1982 and 1985-1987. Not 1991 however, since that episode ended right before Christmas.

For each conflict-year, I therefore need to construct a dummy to indicate whether it is a Christmas conflict or not. I do it in three steps. First, I construct the variable xmas to set the Christmas Day for each year. Second, I set the variable count to 0 if the conflict-year’s ep_end_date occurs before Christmas. Finally, for all cases where ep_end_date is NA, I set count to 1.

df <- here::here("datasets", "ucdp-prio-acd-221.csv") %>%
  read_csv() %>%
  mutate(
    xmas = as.Date(paste0(year, "-12-25")),
    count = ifelse(!(ep_end_date > xmas), 0, 1)
  ) %>%
  replace_na(list(count = 1))

Then it’s a simple matter of counting the active conflicts per year and constructing a bar chart. to illustrate the depressing fact that Christmas 2021 saw an all-time high in active conflicts worldwide.

To end, here’s the song that inspired this post. Have a bad Christmas season everybody.

< section class="quarto-appendix-contents">

Reuse

https://creativecommons.org/licenses/by/4.0/

To leave a comment for the author, please follow the link and comment on their blog: Two Points Make a Line.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.

Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Reuse

Related