Sustainable AI readiness of countries and economies: a hybrid multi-criteria framework integrating compute capacity, energy, carbon, and water constraints — data and code

Data, code, and outputs required to reproduce the weighting, ranking, androbustness results of the manuscript "Sustainable AI Readiness of Countries andEconomies: A Hybrid Multi-Criteria Framework Integrating Compute Capacity,Energy, Carbon, and Water Constraints" (manuscript under review). Eleven indicators across six pillars for 49 countries and economies. Indicatorweights from Shannon entropy and CRITIC; pillar weights from the linearBest-Worst Method (18 of 25 expert responses retained); ranking with MARCOS;cross-validation with TOPSIS and PROMETHEE-II; 1000-run Monte Carlo weightperturbation and six further robustness analyses. Version 1.1.0 adds clustering.py (the script producing the Ward dendrogram andcluster memberships), English-language figures, and an English README. Code MIT, data CC BY 4.0. Third-party sources retain their own licences (seeSOURCES.md). Expert survey data are shared in de-identified form.