A practical flight-phase approach to balanced random sampling

This paper introduces a straightforward method for selecting a balanced random sample from a population. The procedure involves a flight phase, which transforms the vector of inclusion probabilities into one with components close to 0 or 1, followed by a landing phase to complete the selection. We present a novel implementation of the flight phase that leverages linear programming, enabling a highly concise and easily interpretable R code. The method is formally described, implemented in R, and illustrated using real population data. This approach offers a practical, transparent, and reproducible solution to the balanced sampling problem, while establishing a direct link to linear programming techniques.