Figures (Climate Change)
Requirements
These six packages are needed to create the visualizations.
library(janitor) # Cleaning Data
library(tidyverse) # Data Manipulation
library(plm) # Panel Linear Models
library(stargazer) # Table Creation
library(broom) # Tidy data
library(nlme) # Mixed Effects ModelingThis is the final data that we created in the previous script.
p_df <- read.csv("p_df_climate.csv") # Reading in the data
df <- as.data.frame(p_df)
countries <- subset(df, dis == 1) # 1833 country/year occurrences of a disasterLook at the summary statistics
vars_sum <- c(4,6,7,13:16, 18:24) # Sub-setting the variables of interest
stargazer(df[,vars_sum], type="text", out= "table.html", title="Table 1",
align = TRUE)##
## Table 1
## ============================================================
## Statistic N Mean St. Dev. Min Max
## ------------------------------------------------------------
## reg_corr 3,740 0.500 0.307 0.002 0.971
## gdppc 3,469 15.369 16.655 0.538 100.865
## damages 3,741 344,803.100 4,135,080.000 0 210,000,000
## occurences 3,741 1.153 2.109 0 27
## n_floods 3,741 0.914 1.677 0 25
## n_drought 3,741 0.091 0.300 0 3
## oecd 3,741 0.196 0.397 0 1
## l_dam 3,741 2.297 4.750 0.000 19.163
## l_aff 3,741 4.726 5.548 0.000 19.663
## l_death 3,741 1.155 1.854 0.000 12.313
## dis 3,741 0.490 0.500 0 1
## flood 3,741 0.441 0.497 0 1
## drought 3,741 0.087 0.282 0 1
## ------------------------------------------------------------country_counts <- table(df$country) # Which countries had less than 21 values
less_than_21 <- subset(country_counts, country_counts < 21)
head(less_than_21) # These countries have been recently established (checks out)##
## Palestine/Gaza South Sudan
## 14 10Figure 1
Histrogram of Political Corruption
fig1 <- df %>%
ggplot(aes(x=corr)) +
geom_histogram(binwidth=.05, fill="steelblue", color="white", alpha=0.9) +
ggtitle("Figure 1", subtitle ="Public Corruption Distribution") +
theme(plot.title = element_text(hjust = .5, face="bold"),
plot.subtitle = element_text(hjust = .5)) +
xlab("Level of Corruption")
fig1
Figure 2
Political Corruption by Year
df_year <- df %>% group_by(year) %>% summarise_at(vars(corr,occurences,
n_floods,n_earth,n_drought),
list(mean = mean), na.rm= TRUE)
melt_df <- reshape2::melt(df_year, id.var = "year")
vars2 <- "corr_mean"
fig2_df <- melt_df %>% filter(variable %in% vars2)
fig2 <- ggplot(fig2_df, aes(x = year, group=1)) +
geom_line(aes(color = variable, y= value),lwd= 1.5, show.legend = FALSE) +
ggtitle("Figure 2", subtitle = "Public Sector Corruption Yearly Mean") +
xlab("Year") + labs(colour="Variable") + ylim(.4,.6) +
scale_colour_manual(labels = c("Corr"), values = c("brown")) +
ylab("Corruption Level") +
theme(plot.title = element_text(hjust = .5, face="bold"),
plot.subtitle = element_text(hjust = .5))
fig2
Figure 3
Pie Chart Showing Percentage of Each Disaster.
df_pie <- data.frame(event = c("Floods", "Droughts", "Earthquakes"),
count = c(3418, 339, 557))
# calculate percentages
df_pie$percent <- df_pie$count / sum(df_pie$count) * 100
# format labels
df_pie$label <- paste("", df_pie$count, "\n(", round(df_pie$percent, 1), "%)", sep = "")
fig3 <- ggplot(df_pie, aes(x = "", y = count, fill = event)) +
geom_bar(stat = "identity", color = "white") +
coord_polar(theta = "y") +
theme_void() +
labs(title = "Figure 3",
caption = paste("Total:", sum(df_pie$count)),
subtitle = "Natural Disaster Events 2000-2020") +
scale_fill_discrete(name = "Event", labels = paste("", df_pie$event)) +
geom_text(aes(label = label), position = position_stack(vjust = 0.5)) +
theme(plot.title = element_text(hjust = .5, face="bold"),
plot.subtitle = element_text(hjust = .5)) +
scale_fill_brewer(palette="Set1")
fig3