Multivariate Stratified Sampling of Continuous Variables

This repository contains work on multivariate stratified sampling of continuous variables using clustering-based stratification and adapted-Neyman allocation.

Repository Structure

• Local_testing/ — Scripts for local testing of specific seeds and methods.

• multivar_datasets/ — Simulated and applied datasets used for the analysis.

• Kmedoids/ — Run files for k-medoids clustering, with both Neyman-adapted and proportional allocation.

• bb26/ — Benchmark method run files (SamplingStrata) for dataset × strata × sample size combinations.

• CV/ — Minimum CV supplied to SamplingStrata for each dataset × strata × sample size combination.