Reproducibility package for “Universal Service, Unequal Reliability? Territorial Vulnerability and Electricity Service Continuity in a Regulated European Power System”
Description
This dataset provides the complete reproducibility package accompanying the manuscript “Universal Service, Unequal Reliability? Territorial Vulnerability and Electricity Service Continuity in a Regulated European Power System”. The study examines municipality-level electricity service continuity across all 278 municipalities of mainland Portugal during 2014–2025 and its association with a pre-specified 2011 Territorial Vulnerability Index. The package contains the frozen analysis-ready dataset, data dictionary, canonical supporting datasets, exact result outputs, spatial and temporal robustness inputs, quality-assurance records, SHA-256 manifests, publication figures, and reference implementations in Python, R and SPSS, together with an Excel/Calc reproduction guide. It also includes a software-agnostic reproduction recipe specifying the transformations, estimands and inferential procedures independently of any single software environment. Source data originate from publicly available official sources including E-REDES, Statistics Portugal (INE), DGEG, DGT/COS, ICNF and ERA5-Land. Large original public-source snapshots are not duplicated where unnecessary; their provenance, retrieval information and verification hashes are documented in the source manifest. No personal or confidential data are included.
Files
Steps to reproduce
Verify the SHA-256 checksum of analysis_ready/analysis_ready_municipality_2014_2025.csv against SHA256SUMS.txt. Follow SOFTWARE_AGNOSTIC_REPRODUCTION_RECIPE.md for the complete definition of variables, transformations, estimands and model sequence. To reproduce the primary M0–M2 OLS point estimates, run scripts/python/01_point_estimates.py, scripts/r/01_point_estimates.R, or follow the SPSS/Excel equivalents supplied in the package. To reproduce the exact CR2/Satterthwaite inference reported in the manuscript, run scripts/r/02_cr2_satterthwaite.R or implement the formula specified in the Supplementary Material. Compare reproduced estimates with results/primary_M0_M2_cr2.csv and use the QA and SHA-256 manifests to verify the remaining frozen outputs.