Data and analysis code for annual maximum daily precipitation in Campina Grande, Brazil: Gumbel estimates, record coverage and uncertainty

Processed datasets and Python analysis code supporting the study "Annual maximum daily precipitation in Campina Grande, Brazil: Gumbel estimates, record coverage and uncertainty". The study evaluates the effects of record completeness, annual-sample selection and sampling uncertainty on Gumbel rainfall quantiles using INMET conventional station 82795, with an exploratory hourly intensity-duration-frequency analysis using station A313. The package contains processed results, analysis and figure-generation scripts, a data dictionary, software requirements, input-file fingerprints and retrieval instructions. The 30 numerical CSV/JSON analysis outputs were reproduced locally and matched the archived results byte for byte. Original bulk INMET exports are not redistributed; obtain those inputs from INMET using the instructions in README.md. Daily records span 1975 to 2025; 2025 is excluded from fitted annual samples. The reference sample contains 18 complete years, and the sensitivity sample contains 26 years with at least 95% coverage. Results and the exploratory IDF relationship are conditional on sample selection and extrapolation, and are not validated hydraulic design values. Processed results and documentation are licensed under CC BY 4.0; source code is licensed under MIT, as specified in the included license files. The research was self-funded by Francisco de Oliveira Mesquita.