Click the Original, Code and Reconstruction tabs to read about the issues and how they were fixed.

Original


Timor-Leste Ministry of Finance. (2019).


Objective

The original data visualisation targeted investors, government officials, and members of the public with interest in the financial allocations in the Timor-Leste Petroleum Fund. The visualisation’s objective is to allow comparison of the targeted benchmark asset allocation of the fund, and compare this with the actual allocations as at year end.

The visualisation chosen had the following three main issues:

  • The target and actual allocations are shown on two separate doughnut charts making it difficult to compare the two sets of values.

  • The visualisation breaks down the Actual Allocations by Fund Manger. This level of breakdown is not present for the benchmark allocations, making comparison between benchmark and actual allocations difficult.

  • As doughnut charts are used, it is difficult to compare allocations using angle and area without reading the numeric values.

Reference

  • Timor-Leste Ministry of Finance. (2019). Timor-Leste Petroleum Fund Annual Report 2018 . Retrieved fromhttps://www.mof.gov.tl/wp-content/uploads/2019/08/2018-annual-report-ENGLISH.pdf

Code

The following code was used to fix the issues identified in the original.

library(ggplot2)
library(dplyr)

# Read source data extracted from report page 11 to CSV.
# Remove Manager, as it is not included in the Benchmark allocation values to compare with.
# Convert allocation to percentage
data <- readr::read_csv("Timor Leste Petroleum1.csv") %>% 
  group_by(AssetClass, BenchmarkActual) %>% 
  summarise(Allocation = sum(Allocation) / 100)


# Factor BenchmarkActual so order Benchmark first
data$BenchmarkActual <- data$BenchmarkActual %>% 
  factor(levels = c("Benchmark", "Actual"), ordered = TRUE)

# Determine AssetClass Factor level based on largest benchmark Allocation
AssetClassLevels <- data%>% 
  filter(BenchmarkActual == "Benchmark") %>% 
  group_by(AssetClass) %>% 
  summarise(TotalAlloc = sum(Allocation)) %>% 
  arrange(TotalAlloc) %>% 
  pull(AssetClass)

# Factor AssetClass by largest benchmark allocation
data$AssetClass <- data$AssetClass %>% 
  factor(levels = as.list(AssetClassLevels), ordered = TRUE)

p1 <- ggplot(data = data, aes(x = AssetClass, y = Allocation, fill = BenchmarkActual)) + 
  scale_y_continuous(labels = scales::percent_format(accuracy = 1L))+
  geom_bar(stat = "identity",position = position_dodge2(reverse = TRUE)) +
  labs(title = "Asset Allocation vs Benchmarks",
       x = "Asset Class",
       y = "Allocation Percentage") +
  scale_fill_manual('Allocation to:', values = c("steelblue2","turquoise2"))+
  coord_flip()

Data Reference

  • Timor-Leste Ministry of Finance. (2019). Timor-Leste Petroleum Fund Annual Report 2018 . Retrieved fromhttps://www.mof.gov.tl/wp-content/uploads/2019/08/2018-annual-report-ENGLISH.pdf

Reconstruction

The following plot fixes the main issues in the original.