Stacked Barplots USPersonalExpenditure

Note: This article shows an R example of simple stacked and proportions stacked barplots. For more general information, please refer to the following entries: Stacked Barplots and Barplots, Histograms and Boxplots.

In short: In this article, the dataset USExpenditures is visualised using simple and proportions stacked barplots.

The Dataset[edit]

The dataset contains personal expenditures of US citizens (in billions of dollars) in the categories food and tobacco, household operation, medical and health, personal care, and private education for the years 1940, 1945, 1950, 1955 and 1960.

Examining the Dataset[edit]

Therefore, it has five columns for the years and five rows for the categories, which can be inspected using View or head.

USexp<-USPersonalExpenditure
head(USexp)

Output[edit]

> head(USexp)
                      1940   1945  1950 1955  1960
Food and Tobacco    22.200 44.500 59.60 73.2 86.80
Household Operation 10.500 15.500 29.00 36.5 46.20
Medical and Health   3.530  5.760  9.71 14.0 21.10
Personal Care        1.040  1.980  2.45  3.4  5.40
Private Education    0.341  0.974  1.80  2.6  3.64

Simple stacked barplot[edit]

In order to create aesthetically appealing visualisations, you can resort to packages that include nice colour combinations. Here, we are going to use the Wes Anderson package, which has to be installed first.

install.packages("wesanderson")
library("wesanderson")

In this example, the margin around the graphic on the right is extended to fit the legend. Then the stacked barplot is created and subsequently a legend is added.

par(mar = c(5.1, 4.1, 4.1, 10), xpd = TRUE)  #extend margin

barplot(USexp,
        col = wes_palette("Darjeeling2"),
        border = "white", 
        space = 0.04, 
        font.axis = 2,
        xlab = "group",
        ylim = c(0, 200))

#create legend on the right side of the plot
legend("topright", inset = c(-0.33, 0),
       legend = rownames(USexp),
       fill = wes_palette("Darjeeling2"),
       border = "white",
       bty = "n")

par(mar = c(5.1, 4.1, 4.1, 2.1)) #reset margin

Output[edit]

US Personal Expenditure in 1940, 1945, 1950, 1955 and 1960

Total personal expenditure rose over the 20 years displayed here. To see the proportions of the amount spent on the different categories and whether they changed, we can create a percentage plot.

Proportions stacked barplot[edit]

USexp_percentage <- apply(USexp, 2, function(x){ x * 100 / sum(x, na.rm = T) })

par(mar = c(5.1, 4.1, 4.1, 10), xpd = TRUE)  #extend margin

percplot <- barplot(USexp_percentage, col =wes_palette("Moonrise3"), #trying another colour palette
                    border = "white", xlab = "group")

# add percentage values
text(x = rep(percplot, each = nrow(USexp_percentage)),
     y = apply(USexp_percentage, 2, cumsum) - USexp_percentage / 2,
     labels = paste0(round(USexp_percentage, 1), "%"),
     cex = 0.7,
     col = "black")

legend("topright", inset = c(-0.30, 0),
       legend = rownames(USexp),
       fill = wes_palette("Moonrise3"),
       border = "white",
       bty = "n")

par(mar = c(5.1, 4.1, 4.1, 2.1))  #reset margin