Visualizing FDI Pakistan
Salahuddin
17 August, 2020
Initial setup.
setwd("C:/Users/Salahuddin/Desktop/R/SBP")
library(tidyr)
library(dplyr)
library(data.table)
library(ggplot2)Before importing the data lets specify the column types so we can correctly import our data.
In our data we have missing values which are denoted by “-”. We need to specify that these represent missing values by using the na.strings argument. Now lets import the data.
fdi_country <- read.csv("foreign_investment_by_country.csv",
na.strings = "-",
colClasses = col_classes)Lets verify that we have correctly imported our data.
## 'data.frame': 60 obs. of 7 variables:
## $ Country : chr "Argentina" "Australia" "Austria" "Bahamas" ...
## $ Inflow_FY20 : num NA 0.1 3.8 NA 20.1 NA 1.9 NA NA 0.3 ...
## $ Outflow_FY20: num NA NA NA NA NA NA 0 NA NA 0 ...
## $ Net_FY20 : num NA 0.1 3.8 NA 20.1 NA 1.9 NA NA 0.3 ...
## $ Inflow_FY19 : num NA 0.5 7.6 22.9 24.4 NA 4.9 NA 4.4 0.6 ...
## $ Outflow_FY19: num NA 24.6 NA NA 2.6 NA 2 NA NA 0.1 ...
## $ Net_FY19 : num NA -24.1 7.6 22.9 21.9 NA 2.9 NA 4.4 0.5 ...## Country Inflow_FY20 Outflow_FY20 Net_FY20 Inflow_FY19 Outflow_FY19
## 55 United Kingdom 179.9 62.6 117.3 225.0 40.0
## 56 United States 106.9 9.7 97.2 127.5 39.4
## 57 Others 221.2 149.5 71.7 137.6 124.4
## 58 Foreign Private 3285.8 724.6 2561.2 2785.2 1422.8
## 59 Foreign Public NA NA NA NA NA
## 60 Total 3285.8 724.6 2561.2 2785.2 1422.8
## Net_FY19
## 55 185.0
## 56 88.1
## 57 13.2
## 58 1362.4
## 59 NA
## 60 1362.4We don’t need the last three rows since they represent totals. Lets drop those rows.
In the two columns that we are interested, Net_FY20 and Net_FY19, we have missing values. Lets remove them.
# remove rows with missing data from Net_FY20 and Net_FY19
fdi_country <- fdi_country %>%
drop_na(Net_FY20, Net_FY19)We want to work with a few countries since most countries have insignificant net investment values. Lets filter our data.
# filter the data to see which countries have low net investment values
fdi_country %>%
arrange(Net_FY20) %>%
head()## Country Inflow_FY20 Outflow_FY20 Net_FY20 Inflow_FY19 Outflow_FY19
## 1 U.A.E 137.6 181.8 -44.2 160.2 56.4
## 2 Finland 4.0 9.9 -6.0 0.6 3.1
## 3 France 2.9 4.1 -1.2 12.4 NA
## 4 Philippines NA 0.2 -0.2 2.8 0.3
## 5 Indonesia NA 0.1 -0.1 0.1 NA
## 6 Iran NA 0.1 -0.1 0.3 0.4
## Net_FY19
## 1 103.7
## 2 -2.5
## 3 12.4
## 4 2.6
## 5 0.1
## 6 0.0# filter countries with significant net investment values
fdi_country <- fdi_country %>%
filter(Net_FY20 >= 80 | Net_FY20 <= -10)We are only interested in three columns. Lets drop the columns that we don’t need.
# drop columns from the filtered data
drops <- c("Inflow_FY20", "Outflow_FY20",
"Inflow_FY19", "Outflow_FY19")
# drop columns specified above
fdi_country <- fdi_country[ , !(names(fdi_country) %in% drops)]
head(fdi_country)## Country Net_FY20 Net_FY19
## 1 China 844.1 130.8
## 2 Hongkong 190.7 171.0
## 3 Malta 222.2 -139.9
## 4 Netherlands 133.2 69.0
## 5 Norway 401.9 115.8
## 6 U.A.E -44.2 103.7Since we’d be creating a side by side bar chart, we need to reshape our data. We will create two new columns, FY and Net_FDI.
# melt the data and re-stack the columns
fdi_country <- melt(fdi_country, measure.vars = c("Net_FY20", "Net_FY19"),
variable.name = "FY", value.name = "Net_FDI")
head(fdi_country)## Country FY Net_FDI
## 1 China Net_FY20 844.1
## 2 Hongkong Net_FY20 190.7
## 3 Malta Net_FY20 222.2
## 4 Netherlands Net_FY20 133.2
## 5 Norway Net_FY20 401.9
## 6 U.A.E Net_FY20 -44.2We are now ready to start plotting our graph.
p1 <- ggplot(fdi_country, aes(x = Country, y = Net_FDI,
fill = FY)) +
geom_bar(stat = "identity", width = .7, position = "dodge")
p1
Lets add a horizontal line at \(y = 0\).
R automatically assigns colors to the bars and also assigns legend labels. But we can manually specify the colors and the labels.
p3 <- p2 + scale_fill_manual(labels = c("Net Investment FY20",
"Net Investment FY19"),
values = c("#006991","#0f9fd5"))
p3
We can see that the country names are overlapping. We can rename that countries to make them shorter but here I want the bars horizontally. Lets fix that by flipping the coordinates.
At this point, lets add some labels to our graph.
p5 <- p4 + ggtitle ("Net Foreign Investment in Pakistan by Countries") +
labs(y = "Net FDI\n", subtitle = "Net Foreign Investment, in Million US $") +
labs(caption = "\nSource: State Bank of Pakistan \n")
p5
Now we can start customizing our graph. We can choose an in-built theme by passing the theme argument but I want to build my own theme. Lets see how it works by removing the axis ticks and the background color.
p6 <- p5 + theme(panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.background = element_blank(),
line = element_blank())
p6
I want to change the position of the y axis and expand the limits. I also want to add intervals of 100.
p7 <- p6 + expand_limits(y = c(-200, 1000)) +
scale_y_discrete(expand = c(0, 8), position = "right",
breaks = c(-100, 0, 100, 200, 300, 400, 500, 600, 700, 800),
lim = c(-100, 0, 100, 200, 300, 400, 500, 600, 700, 800))## Warning: Continuous limits supplied to discrete scale.
## Did you mean `limits = factor(...)` or `scale_*_continuous()`?
Lets add the axis line and also add some background color.
p8 <- p7 + theme(axis.line.x = element_line(),
rect = element_rect(fill = "#d5e7e8"),
panel.background = element_rect(fill = "#d5e7e8"),
panel.grid.major.y = element_line(color = "white"))
p8
I want to change the position of the legend and move it to top left.
p9 <- p8 + theme(legend.title = element_blank(),
legend.position = "top",
legend.justification = c(0, 0),
legend.direction = "horizontal")
p9
The graph title and sub-title can also be customized.
p10 <- p9 + theme(plot.title = element_text(colour = "black", size = 14,
face = "bold"),
plot.subtitle = element_text(colour = "black", size = 12))
p10
Finally, we can add data labels,
# add labels
p11 <- p10 + geom_text(aes(y = max(Net_FDI) + 100, label = Net_FDI),
position = position_dodge(width = 1),
size = 3.2)
p11
We can position the labels on the bars but I wanted to keep them on the right.
That’s it!