Rows: 355 Columns: 42
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (12): Country, ISO2, ISO3, Indicator, Unit, Source, CTS_Code, CTS_Name, ...
dbl (30): ObjectId, F1985, F1986, F1987, F1990, F1991, F1992, F1993, F1994, ...
ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
# we want to compare these two indicatorsindicators_we_want <-c("Green Bond Issuances by Country", "Sovereign Green Bond Issuances")green_debt_subset <- green_debt |># from the janitor package -- makes variables snake_case so they are easier to work withclean_names() |># filter for the vector of indicators we defined abovefilter(indicator %in% indicators_we_want) |># "f\\d{4}" is a regular expression (regex) that searches for all columns that are f + four digits.# Ask ChatGPT to explain this to you.select(country, iso3, indicator, matches("f\\d{4}")) library(countrycode)# Assuming your country column is named 'country' and it contains country names or ISO codes.# This will create a new column in your data frame called 'region' with the corresponding region names.green_debt_subset$region <-countrycode(green_debt_subset$country, "country.name", "region")green_bonds_tidy <- green_debt_subset |>pivot_longer(# select all coluns with f + 4 numbers cols =matches("f\\d{4}"),# change from default ("names")names_to ="year",# same with the valuesvalues_to ="issuance_bn_usd",# readr::parse_number is a handy function that changes the character string # "f2222" into the number 2222. Very useful! names_transform = readr::parse_number,# green bonds are new-ish, we can drop all those NA values in the 80s and 90s for now.values_drop_na =TRUE )# Load necessary librarieslibrary(tidyverse)library(ggplot2)# Read the datasetgreen_bonds_data <-green_bonds_tidy %>%filter(year >=2012) # Calculate cumulative issuance by regiongreen_bonds_cumulative <- green_bonds_data %>%group_by(region, year) %>%summarise(annual_issuance =sum(issuance_bn_usd)) %>%arrange(region, year) %>%group_by(region) %>%mutate(cumulative_issuance =cumsum(annual_issuance))
`summarise()` has grouped output by 'region'. You can override using the
`.groups` argument.
# Reshape data if necessary using pivot_longer and pivot_wider# Create the visualizationggplot(green_bonds_cumulative, aes(x = year, y = cumulative_issuance, color = region)) +geom_line() +geom_point() +scale_x_continuous(breaks =seq(min(green_bonds_cumulative$year), max(green_bonds_cumulative$year), by =1)) +scale_y_continuous(labels = scales::dollar_format(suffix ="B", scale =1)) +labs(title ="Cumulative Green Bond Issuance by Region",x ="Year",y ="Cumulative Issuance (Billion USD)",color ="Region") +theme_minimal() +theme(legend.position ="bottom",plot.title =element_text(hjust =0.5),axis.text.x =element_text(angle =45, hjust =1))
Homework problem 2:
green_debt_HWP2 <- green_debt |>clean_names() |>filter(type_of_issuer !="Not Applicable") |>pivot_longer(# select all coluns with f + 4 numbers cols =matches("f\\d{4}"),# change from default ("names")names_to ="year",# same with the valuesvalues_to ="issuance_bn_usd",# readr::parse_number is a handy function that changes the character string # "f2222" into the number 2222. Very useful! names_transform = readr::parse_number,# green bonds are new-ish, we can drop all those NA values in the 80s and 90s for now.values_drop_na =TRUE )# Load necessary librarieslibrary(tidyverse)library(ggplot2)# Filter the dataset for years after 2012HWP2_issuer1 <- green_debt_HWP2 %>%filter(year >2012)# Group and summarize the data by type_of_issuer and year issuer_annual_totals <- HWP2_issuer1 %>%group_by(type_of_issuer, year) %>%summarise(issuance_bn_usd =sum(issuance_bn_usd, na.rm =TRUE),.groups ='drop')# Create a data visualizationggplot(issuer_annual_totals, aes(x = year, y = issuance_bn_usd, color = type_of_issuer)) +geom_line() +geom_point() +labs(title ="Annual Green Bond Issuance by Type of Issuer (Post-2012)",x ="Year",y ="Issuance (Billion USD)",color ="Type of Issuer") +theme_minimal() +theme(legend.position ="bottom",plot.title =element_text(hjust =0.5),axis.text.x =element_text(angle =45, hjust =1))
# Assuming green_debt_HWP2 has been read into R# First, filter the dataset for years after 2012HWP2_issuer2_filtered <- green_debt_HWP2 %>%filter(year >=2012) %>%group_by(year) %>%mutate(total_annual_issuance =sum(issuance_bn_usd, na.rm =TRUE)) %>%ungroup() %>%group_by(type_of_issuer, year) %>%summarise(issuance_bn_usd =sum(issuance_bn_usd, na.rm =TRUE),total_annual_issuance =first(total_annual_issuance),market_share = issuance_bn_usd /first(total_annual_issuance),.groups ='drop' ) %>%ungroup() %>%group_by(type_of_issuer) %>%mutate(year_over_year_growth = (issuance_bn_usd /lag(issuance_bn_usd) -1) *100) %>%ungroup()# Separate plot for Market Share by Issuer Typeggplot(HWP2_issuer2_filtered, aes(x = year, y = market_share, fill = type_of_issuer)) +geom_bar(stat ="identity", position ="dodge") +scale_x_continuous(breaks =function(x) unique(HWP2_issuer2_filtered$year)) +# Ensure x-axis has integer yearslabs(title ="Market Share by Issuer Type (Post-2012)",x ="Year",y ="Market Share",fill ="Type of Issuer") +theme_minimal() +theme(legend.position ="bottom")
Homework problem 3:
Q1：What are green bond proceeds used for?
green_debt_HWP3.1<- green_debt |>clean_names() |>filter(use_of_proceed !="Not Applicable") |>pivot_longer(# select all coluns with f + 4 numbers cols =matches("f\\d{4}"),# change from default ("names")names_to ="year",# same with the valuesvalues_to ="issuance_bn_usd",# readr::parse_number is a handy function that changes the character string # "f2222" into the number 2222. Very useful! names_transform = readr::parse_number,# green bonds are new-ish, we can drop all those NA values in the 80s and 90s for now.values_drop_na =TRUE)# Load necessary librarieslibrary(tidyverse)library(ggplot2)# Filter out data to focus on the absolute dollar amounts for the use of proceedsgreen_debt_HWP3.2<- green_debt_HWP3.1%>%filter(year ==2022& unit =="Billion US Dollars")# Aggregate issuance volumes by use_of_proceedissuance_summary_2022 <- green_debt_HWP3.2%>%group_by(use_of_proceed) %>%summarise(total_issuance =sum(issuance_bn_usd, na.rm =TRUE)) %>%ungroup()# Identify the top 10 proceeds based on total issuancetop_issuance_2022 <- issuance_summary_2022 %>%top_n(10, total_issuance)# Plotting the top 10 proceedsggplot(top_issuance_2022, aes(x =reorder(use_of_proceed, total_issuance), y = total_issuance, fill = use_of_proceed)) +geom_bar(stat ="identity") +coord_flip() +# Flip coordinates to make it a horizontal bar chartscale_fill_viridis_d() +# Use a discrete viridis color scalelabs(title ="Top 10 Green Bond Proceeds in 2022 by Cumulative Issuance Volume",x ="Total Issuance (Billion USD)", y ="Use of Proceeds") +theme_minimal() +theme(legend.title =element_blank(), # Hide the legend titlelegend.position ="none") # Hide the legend as the information is already on the y-axis
# Show the plotggsave("top_10_proceeds_2022.png", width =10, height =8) # Save the plot as a PNG file
Q2：What do we know about the currency of issuance? Is that changing over time?
green_debt_HWP3.3<- green_debt |>clean_names() |>filter(principal_currency !="Not Applicable") |>pivot_longer(# select all coluns with f + 4 numbers cols =matches("f\\d{4}"),# change from default ("names")names_to ="year",# same with the valuesvalues_to ="issuance_bn_usd",# readr::parse_number is a handy function that changes the character string # "f2222" into the number 2222. Very useful! names_transform = readr::parse_number,# green bonds are new-ish, we can drop all those NA values in the 80s and 90s for now.values_drop_na =TRUE)library(ggplot2)library(dplyr)library(readr)# Load the datasetgreen_debt_currency <- green_debt_HWP3.3# Summarize issuance volume by currency for 2022currency_summary <- green_debt_currency %>%filter(year ==2022, unit =="Billion US Dollars") %>%group_by(principal_currency) %>%summarise(total_issuance =sum(issuance_bn_usd, na.rm =TRUE)) %>%mutate(proportion = total_issuance /sum(total_issuance)) %>%arrange(desc(total_issuance)) %>%slice_max(order_by = total_issuance, n =10)# Plotting the top 10 currenciesggplot(currency_summary, aes(x =reorder(principal_currency, total_issuance), y = total_issuance)) +geom_bar(stat ="identity", fill ="steelblue") +geom_text(aes(label = scales::percent(proportion)), vjust =-0.5, size =3.5) +labs(title ="Top 10 Currencies for Green Bond Issuance Volume in 2022",subtitle ="Volume in Billion USD and Proportion of Total",x ="Currency", y ="Issuance Volume (Billion USD)") +theme_minimal() +theme(axis.text.x =element_text(angle =45, hjust =1))
