Cluster analysis applied to fuel price data in Campina Grande

This work uses cluster analysis to analyze similarities in the prices of regular gasoline and ethanol in Campina Grande-PB, Brazil. This multivariate statistical technique groups similar elements into distinct clusters using the non-hierarchical method (k-means). The analysis reports three clusters: cluster 1 has the lowest prices, while cluster 3 has the highest. We report descriptive statistics and a discussion based on the empirical local context.