Skip to content
Quantoros

What does Latin hypercube sampling guarantee in each one-dimensional marginal?

MathematicsLevel: IntermediateSimulation design

What does Latin hypercube sampling guarantee in each one-dimensional marginal?

Hints
  • Check each coordinate separately.
  • Marginal coverage is not joint coverage.

Answer

With n points, each of n equal-probability intervals gets one point. This stratifies marginals but does not guarantee all multidimensional cells are filled.

Source and licence: SciPy — quasi-Monte Carlo designs · BSD-3-Clause · answer: Quantoros editorial

Related questions