ABSTRACT
This chapter introduces the Latin hypercube sampling method for selecting representative variable annuity policies. In statistical sampling, a Latin square is a square grid of sample positions from a two-dimensional space such that there is only one sample in each row and each column. A Latin hypercube is the generalization of a Latin square to a high dimensional space such that each axis-aligned hyperplane contains only a single sample. The quality of a Latin hypercube can also be measured by the minimum distances between its sample positions. The function is named calDist and has three arguments. The first argument is a matrix, which represents a Latin hypercube sample created by the function mylhs. The second and the third arguments are the vectors of lower and upper bounds of the continuous variables.
