Replication package for: Determinants of Curtailment in Brazil: Renewable Pressure, Absorption Capacity and Margins of Adjustment

Data and code that regenerate every table, figure and number of the paper Determinants of Curtailment in Brazil: Renewable Pressure, Absorption Capacity and Margins of Adjustment, and a script that checks each regenerated number against the manuscript. Data: an audited submarket-day panel of renewable constrained-off in Brazil's National Interconnected System (4 submarkets, 1 April 2024 to 5 May 2026, 3,060 observations), with the curtailment rate and renewable pressure built from potential (pre-curtailment) generation, and public hourly extracts (ONS, CCEE) from which a 68,672-observation submarket-hour panel is rebuilt. Methods: hurdle model (Firth-penalised logit for occurrence, beta regression for intensity) with a moving-block bootstrap; linear fixed-effects benchmarks with Driscoll-Kraay standard errors; functional-form and threshold tests; hourly zero/one-inflated specifications; SUR, VARX, Granger and cointegration tests. All estimates are conditional associations. Use: pip install -r requirements.txt and python run_all.py (several hours; --quick for a smoke test). The reference run and a line-by-line verification report against the manuscript are included. Licences: data, tables, figures and documentation under CC BY 4.0; code under the MIT License (LICENSE-CODE). Version 1.1.0 (November 2026): adds s06b_share_bootstrap.py, which reports the two terms of the decomposition of the average marginal effect of renewable pressure on the unconditional mean (Section 4.2, Table 2) and their 14-day block-bootstrap distribution, reproducing the draws of s06; documents the threshold bootstrap of Appendix A.2 (grid search repeated in every resample, fixed-design residual blocks of whole days); 135 printed numbers are verified, of which all 131 regenerable ones match. Results of version 1.0.0 are unchanged.