Replication package for "Testing the Aggregation Rule of Regional Composite Indices with Few Regions: A Procedure, a Python Implementation and an Application to Uzbekistan"

Replication package for the article “Testing the Aggregation Rule of Regional Composite Indices with Few Regions: A Procedure, a Python Implementation and an Application to Uzbekistan”. It contains the Python module regagg 1.0.2 (MIT licence) with its test suite (16 tests), the data extracts and the scripts that reproduce every table and figure of the paper and of the online supplement. The module tests the aggregation rule of a regional composite index against an outcome when regions are few. It profiles a weighted generalised mean, tests averaging, bottleneck and dispersion-penalty rules with a restricted wild cluster bootstrap (Webb weights, Holm adjustment, optional fast double bootstrap), runs a relevance test and splits each test statistic into a weights part and a rule part. Data: regional series for the 14 regions of Uzbekistan, 2014–2025, from the SDMX indicator portal of the Statistics Agency under the President of the Republic of Uzbekistan (https://stat.uz). Usage: pip install . ; pytest tests/ ; python run_all.py (about 8–10 hours on one core). example_synthetic.py runs in a few seconds. See README.md for the script-to-table map.