Cluster Analysis Application

The application performs a Cluster Analysis of the 28 European Union Countries.

• The analysis examines 3 economic indicators (year 2014):
  1. Gross Domestic Product (Euro per capita)
  2. Government Debt (Percentage of Gross Domestic Product)
  3. Unemployment Rate.
• The dataset of the economic indicators is obtained from: Eurostat
• The indicators come from 3 distinct datasets which I cleaned and assembled together.
  

The server.R and ui.R files and the associated supporting documentation are in the following GitHub repository: https://github.com/GValentini/Data_Products.git

The application is available here: https://gvalentini.shinyapps.io/ClustersEur-App/

• The user chooses:
  1. the economic indicators to plot on the 2 axes
  2. the number of centres
     
• The points plotted represent the ordered pairs of the chosen indicators.
• We have one point for each Country.
• The K-Means method creates a partition of the points into k groups (called clusters) such that the sum of squares from points to the assigned cluster centres is minimized.
• Each cluster is identified by a distinct color.
• The cross-shaped symbol represents the cluster centre.
• More details about K-Means Clustering Analysis here: https://en.wikipedia.org/wiki/K-means_clustering

eurdf <- readRDS("eurodata.rds")
result <- kmeans(eurdf[, 4:5], centers = 4)
par(mfrow = c(2, 1), pin = c(7, 2), mar = c(4, 4, 0, 2))
plot(eurdf[, 4:5], col = result$cluster, pch = 19, cex = 1.8); grid(lty = 1)
text(eurdf[, 4:5], labels = eurdf[,2], pos = 3, cex = 0.8, col = result$cluster)
points(result$centers, col = 1:4, pch = 3, cex = 3, lwd = 3)