Origin of crude oil imported to China

Daily Graph Series

Economist-Style Plot using R

The graph is motivated by Economist’s plot and can be replicated by using R as follows:

R Codes for Data Cleaning and Visualization

In this post I use uspopage data set from gcookbook package for illustrative purpose.

rm(list = ls())
library(tidyverse)
library(gcookbook)
library(scales)


my_colors <- c("#A8A9AD", "#5C92A1", "#6ED0F5", "#E7D3BA", "#822813", "#ED9176", "#EF5B3E")
brgColor <- "#CDDEE6"
my_font <- "Ubuntu Condensed"

uspopage %>% 
  filter(AgeGroup != "<5") %>% 
  filter(Year >= 1960) %>% 
  filter(Year <= 2000) -> df_mini

AgeGroup <- df_mini$AgeGroup %>% unique()
levels <- c("Other", "Russia", "Angola", "Other Middle East", "Iraq", "Iran", "Saudi Arabia")

df_mini %>% 
  mutate(Country = case_when(AgeGroup == AgeGroup[1] ~ levels[1], 
                             AgeGroup == AgeGroup[2] ~ levels[2], 
                             AgeGroup == AgeGroup[3] ~ levels[3], 
                             AgeGroup == AgeGroup[4] ~ levels[4], 
                             AgeGroup == AgeGroup[5] ~ levels[5], 
                             AgeGroup == AgeGroup[6] ~ levels[6], 
                             TRUE ~ levels[7])) %>% 
  mutate(Country = factor(Country, levels = levels)) -> dfFake

dfFake %>% 
  filter(Year == 2000) %>% 
  pull(Thousands) %>% 
  sum() ->  max_value

my_breaks <- seq(from = 0, to = max_value, length.out = 7)


# Draft Plot: 

dfFake %>% 
  ggplot(aes(x = Year, y = Thousands, fill = Country)) + 
  geom_area() +
  scale_x_continuous(expand = c(0, 2), limits = c(1960, 2010)) +
  scale_y_continuous(expand = c(0, 1000), breaks = my_breaks, position = "right", 
                     limits = c(0, 270000), labels = 0:6) + 
  theme(legend.position = "right") + 
  scale_fill_manual(values = my_colors)

dfFake %>% 
  group_by(Year) %>% 
  summarise(Thousands = sum(Thousands)) %>% 
  ungroup() -> df_BlackLine


# Function creates data frame for white lines: 

df_whiteLine <- function(Country_selected) {
  
  dfFake %>% 
    filter(!Country %in% Country_selected) %>% 
    group_by(Year) %>% 
    summarise(Thousands = sum(Thousands)) %>% 
    ungroup() -> df
  
  return(df)
}

# Data frame for ploting text: 

df_BlackLine %>% 
  filter(Year == 2000) -> l0

df_whiteLine(levels[1]) -> l1
df_whiteLine(levels[1:2]) -> l2
df_whiteLine(levels[1:3]) -> l3
df_whiteLine(levels[1:4]) -> l4
df_whiteLine(levels[1:5]) -> l5
df_whiteLine(levels[1:6]) -> l6

l0 %>% 
  bind_rows(l1 %>% filter(Year == 2000)) %>% 
  bind_rows(l2 %>% filter(Year == 2000)) %>% 
  bind_rows(l3 %>% filter(Year == 2000)) %>% 
  bind_rows(l4 %>% filter(Year == 2000)) %>% 
  bind_rows(l5 %>% filter(Year == 2000)) %>% 
  bind_rows(l6 %>% filter(Year == 2000)) -> df1

df1 %>% 
  mutate(my_lead = lead(Thousands, 1L)) %>% 
  mutate(my_lead = replace_na(my_lead, 0)) %>% 
  mutate(pos = 0.5*(Thousands + my_lead)) -> df_text

my_label <- case_when(str_detect(levels, "Middle") ~ "Other\nMiddle East", TRUE ~ levels)
  
# Draft plot: 

ggplot() + 
  geom_area(data = dfFake, aes(x = Year, y = Thousands, fill = Country), show.legend = FALSE) + 
  theme_minimal(base_family = my_font) + 
  geom_line(data = df_BlackLine, aes(x = Year, y = Thousands), size = 0.8) + 
  geom_line(data = l1, aes(x = Year, y = Thousands), size = 0.8, color = "white") + 
  geom_line(data = l2, aes(x = Year, y = Thousands), size = 0.8, color = "white") + 
  geom_line(data = l3, aes(x = Year, y = Thousands), size = 0.8, color = "white") + 
  geom_line(data = l4, aes(x = Year, y = Thousands), size = 0.8, color = "white") + 
  geom_line(data = l5, aes(x = Year, y = Thousands), size = 0.8, color = "white") + 
  geom_line(data = l6, aes(x = Year, y = Thousands), size = 0.8, color = "white") + 
  scale_x_continuous(expand = c(0, 0.5), limits = c(1960, 2006)) +
  scale_y_continuous(expand = c(0, 1000), breaks = my_breaks, position = "right",
                     limits = c(0, 270000), labels = 0:6) +
  theme(legend.position = "right") + 
  scale_fill_manual(values = my_colors) + 
  geom_point(data = df_text, aes(x = Year - 1, y = pos), color = "grey20") + 
  geom_text(data = df_text, aes(x = Year + 1.5, y = pos, label = my_label), 
            hjust = 0, family = my_font, size = 5.5, color = "grey20") + 
  geom_segment(data = df_text, aes(x = Year - 1, xend = Year + 1.3, y = pos, yend = pos), color = "grey20") + 
  theme(panel.grid.major.x = element_blank()) + 
  theme(panel.grid.minor.x = element_blank()) + 
  theme(panel.grid.minor.y = element_blank()) + 
  theme(plot.background = element_rect(fill = brgColor, color = NA)) + 
  theme(panel.background = element_rect(fill = brgColor, color = NA)) + 
  theme(axis.title = element_blank()) + 
  theme(plot.margin = unit(rep(0.8, 4), "cm")) + 
  labs(title = "Growing Dependence", 
       subtitle = "Chinese crude-oil import by country, m b/d", 
       caption = "Source: China Customs") + 
  theme(plot.title = element_text(size = 24)) + 
  theme(plot.subtitle = element_text(size = 16, color = "grey20")) + 
  theme(plot.caption = element_text(size = 13, color = "grey20")) + 
  theme(axis.text = element_text(size = 16))

library(grid)
grid.rect(x = 0.018, y = 0.93, hjust = 1, vjust = 0, gp = gpar(fill = "#e5001c", lwd = 0))