Full Analysis Report

Basic EDA

# Imports
library(refugees) # UNHCR's refugees R package
library(tidyverse)
library(patchwork) # To combine plots
library(grid) # For unit() function to set margin dimensions
library(ggtext) # For Markdown/HTML formatting in text elements

# Load Data
population <- readr::read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2023/2023-08-22/population.csv") %>%
  mutate(`refugees (M)` = refugees / 1e6) # convert to millions


# Filter regions
selected_regions <- population %>%
  filter(coo_iso %in% c("SYR", "AFG", "MMR", "YEM", "SDN", "COD"))

# Check Data types
glimpse(selected_regions)

Rows: 6,401
Columns: 17
$ year              <dbl> 2010, 2010, 2010, 2010, 2010, 2010, 2010, 2010, 2010…
$ coo_name          <chr> "Afghanistan", "Dem. Rep. of the Congo", "Sudan", "D…
$ coo               <chr> "AFG", "COD", "SUD", "COD", "SUD", "COD", "SUD", "SY…
$ coo_iso           <chr> "AFG", "COD", "SDN", "COD", "SDN", "COD", "SDN", "SY…
$ coa_name          <chr> "Afghanistan", "Algeria", "Algeria", "Angola", "Ango…
$ coa               <chr> "AFG", "ALG", "ALG", "ANG", "ANG", "ARE", "ARE", "AR…
$ coa_iso           <chr> "AFG", "DZA", "DZA", "AGO", "AGO", "EGY", "EGY", "EG…
$ refugees          <dbl> 0, 66, 0, 13648, 159, 24, 10035, 8, 205, 6, 5, 5518,…
$ asylum_seekers    <dbl> 0, 26, 20, 1508, 212, 14, 12476, 0, 9, 0, 0, 68, 5, …
$ returned_refugees <dbl> 0, 0, 0, 0, 0, 0, 251, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ idps              <dbl> 351907, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ returned_idps     <dbl> 3366, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
$ stateless         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
$ ooc               <dbl> 838250, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ oip               <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …
$ hst               <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …
$ `refugees (M)`    <dbl> 0.000000, 0.000066, 0.000000, 0.013648, 0.000159, 0.…

selected_regions$coo_name <- as.factor(selected_regions$coo_name) # Convert to factor

conflict_regions <- selected_regions %>%
  mutate(year = as.numeric(year)) %>% # Ensure year is numeric
  group_by(year, coo_name) %>% # Group by year and country of origin
  summarize(total_refugees = sum(`refugees (M)`))

Custom Theme

All visualizations in this analysis follow a custom theme to maintain a consistent style across all plots.

The theme is set globally via .Rprofile.

# Custom Color Palette
region_colors <- c(
  "Syrian Arab Rep." = "tomato2",
  "Afghanistan" = "palevioletred1",
  "Myanmar" = "turquoise1",
  "Sudan" = "orange",
  "Dem. Rep. of the Congo" = "saddlebrown",
  "Yemen" = "olivedrab2"
)

Visualizing Refugee Trends Over Time by Country

# Define the desired order of regions
desired_order <- c(
  "Syrian Arab Rep.", "Afghanistan", "Myanmar", "Sudan",
  "Dem. Rep. of the Congo", "Yemen"
)
# Reorder the coo_name column
conflict_regions$coo_name <- factor(conflict_regions$coo_name, levels = desired_order)

# Generate Line Chart with Facets
ggplot(conflict_regions, aes(
  x = year, y = total_refugees,
  colour = coo_name
)) +
  geom_line(linewidth = 0.75) +
  scale_color_manual(
    values = region_colors
  ) +
  scale_x_continuous(breaks = seq(2010, 2022, by = 4)) +
  labs(
    title = "<b>Refugees (Millions)</b> Over Time by Country",
    x = NULL,
    y = NULL,
    color = "Country of Origin"
  ) +
  facet_wrap(~coo_name, scales = "free_x") + # Independent scaling of x-axis
  theme(
    strip.text = element_text(face = "bold"), # Make facet labels bold
    strip.background = element_blank(),
    panel.spacing = unit(0.85, "cm"),
    plot.title = ggtext::element_markdown(size = 12, hjust = 0.1) # Enable Markdown for the title
  )

Comparing Refugee Trends with Different Y-axis Scaling

This comparison highlights how the choice of scaling can influence the visual interpretation

# List of regions you want to compare
region_names <- c("Myanmar", "Sudan", "Dem. Rep. of the Congo", "Yemen")

# Loop through each region, create facet mini plot and individual plot, then combine them
for (region in region_names) {
  # Filter data for the region
  region_data <- conflict_regions %>%
    filter(coo_name == region)

  # Mini plot from facet chart (fixed scaling)
  facet_plot <- ggplot(region_data, aes(x = year, y = total_refugees, colour = coo_name)) +
    geom_line(linewidth = 0.7) +
    scale_color_manual(
      values = region_colors
    ) +
    scale_x_continuous(breaks = seq(2010, 2022, by = 4)) +
    scale_y_continuous(limits = c(0, max(conflict_regions$total_refugees))) +
    labs(
      title = paste(region, "(from Facet Chart)"),
      x = NULL,
      y = "Refugees (in <b>Millions</b>)"
    ) +
    theme(
      plot.title = element_text(hjust = 0.5),
      legend.position = "none",
      plot.margin = unit(c(0.9, 0.9, 0.9, 0.9), "cm"), # Add 1 cm of padding around the plot
      axis.title.y = ggtext::element_markdown() # Enable Markdown in y-axis title
    )

  # Independent plot (auto-scaled)
  individual_plot <- ggplot(region_data, aes(x = year, y = total_refugees, colour = coo_name)) +
    geom_line(linewidth = 0.7) +
    scale_color_manual(
      values =
        region_colors
    ) +
    scale_x_continuous(breaks = seq(2010, 2022, by = 4)) +
    labs(
      title = paste(region, "(Individual Chart)"),
      x = NULL,
      y = "Refugees (<b>Thousands</b>)"
    ) +
    theme(
      plot.title = element_text(hjust = 0.5),
      legend.position = "none",
      plot.margin = unit(c(0.9, 0.9, 0.9, 0.9), "cm"),
      axis.title.y = ggtext::element_markdown()
    )

  # Combine the two plots side by side using the patchwork
  combined_plot <- facet_plot + individual_plot +
    plot_annotation(
      title = paste("Facet vs Individual Scaling"),
      theme = theme(plot.title = element_text(size = 14, hjust = 0.5))
    )

  # Add extra padding around the combined plot
  combined_plot <- combined_plot & theme(plot.margin = unit(c(1.5, 1.5, 1.5, 1.5), "cm"))

  # Save the combined plot
  ggsave(
    filename = paste0(region, "_comparison_plot.png"),
    plot = combined_plot,
    width = 12,
    height = 6,
    dpi = 300,
    bg = "white"
  )
}

Combined plots (Facet vs Individual Scaling)