Reproducibility package: Typhoid fever in the Republic of Korea, 2001–2024 — long-term decline, residual annual periodicity, and southeastern coastal spatial clustering

This Zenodo deposit is the reproducibility package accompanying the manuscript: Lee G and Kim S. Typhoid fever in the Republic of Korea, 2001–2024: long-term decline, residual annual periodicity and southeastern coastal spatial clustering in 24 years of nationwide surveillance. Submitted to Foodborne Pathogens and Disease (Sage Journals, R2 revision). Contents:- /tables/ — 14 source CSV files underlying main and supplementary tables (annual rates, age-specific Mann–Kendall, Sen slope confidence intervals, wavelet stability, breakpoint bootstrap, LISA FDR, Empirical Bayes sigungu shrinkage, 12-weighting Moran's I sensitivity, n=1,000 power simulations).- /figures/ — 2 main composite figures (Figure 1: 6 panels; Figure 2: 5 panels) and 4 supplementary figures (S1 STL decomposition, S2 elderly share, S3 NB GLM fit overlay, S4 sigungu animated MP4 + interactive HTML).- /scripts/ — 11 Python and JavaScript analysis and figure-rendering scripts (typhoid_r2_analyses.py, recompute_weights.py, rerun_power_n1000.py, rebuild_combined_figs.py, build_sigungu_mp4.py, etc.).- /docs/ — manuscript, supplementary material, cover letter (final R2 versions). Data sources:- Korea Disease Control and Prevention Agency (KDCA) Infectious Disease Portal (https://dportal.kdca.go.kr/) — typhoid notifications, weekly × sigungu, 2001–2024.- Korean Statistical Information Service (KOSIS) — annual mid-year resident registration population by sigungu, 2001–2024. Methods (summary):Hamed–Rao modified Mann–Kendall test with rank-based Sen slope 95% CIs; negative-binomial GLM (NB2, dispersion α = 0.154); Morlet continuous wavelet transform with 1,000-simulation AR(1) red-noise null; exhaustive BIC breakpoint search with 200-replicate parametric bootstrap; direct age-standardisation to the 2005 Korean Standard Population; global Moran's I (9,999 permutations) under Queen + KNN-2 baseline plus 11 alternative weighting schemes; LISA with Benjamini–Hochberg FDR; Empirical Bayes (Marshall 1991) sigungu shrinkage; period-stratified analysis (2001–2012 vs 2013–2024). Reproducibility:All Monte-Carlo procedures use seed = 42. Python 3.12 with pandas, numpy, statsmodels, scipy, geopandas, plotly, matplotlib, PyMannKendall. License: Creative Commons Attribution 4.0 International (CC-BY-4.0)本Zenodo存档为配套以下稿件的可复现研究包： Lee G与Kim S。《2001-2024年韩国伤寒：全国24年监测中的长期下降、残留年度周期性与东南沿海空间聚集性》，已提交至《食源性病原菌与疾病》（Sage期刊，R2修回稿）。 【内容】 - /tables/：包含支撑主表与补充表的14个源逗号分隔值（Comma-Separated Values, CSV）文件，涵盖年度发病率、分年龄曼-肯德尔（Mann-Kendall）检验、森斜率（Sen slope）置信区间、小波稳定性分析、断点自助法、局部空间关联指标错误发现率（Local Indicators of Spatial Association-False Discovery Rate, LISA FDR）、经验贝叶斯市郡区（sigungu）收缩率、12权重Moran's I敏感性分析、1000次功效模拟。 - /figures/：包含2张主合成图（图1含6个子图，图2含5个子图）与4张补充图（S1：基于Loess的季节-趋势分解（Season-Trend decomposition using Loess, STL）；S2：老年人口占比；S3：负二项广义线性模型（Negative Binomial Generalized Linear Model, NB GLM）拟合叠加图；S4：市郡区动画MP4文件与交互式HTML文件）。 - /scripts/：包含11个Python与JavaScript分析及绘图脚本，包括typhoid_r2_analyses.py、recompute_weights.py、rerun_power_n1000.py、rebuild_combined_figs.py、build_sigungu_mp4.py等。 - /docs/：包含稿件、补充材料与投稿信（最终R2版本）。 【数据来源】 - 韩国疾病控制与预防厅（Korea Disease Control and Prevention Agency, KDCA）传染病门户网站（https://dportal.kdca.go.kr/）：2001-2024年按市郡区（sigungu）划分的周度伤寒通报数据。 - 韩国统计信息服务（Korean Statistical Information Service, KOSIS）：2001-2024年按市郡区（sigungu）划分的年度年中常住人口登记数据。 【方法（摘要）】 采用带秩的森斜率95%置信区间的Hamed-Rao修正曼-肯德尔（Mann-Kendall）检验；负二项广义线性模型（NB2，离散度α=0.154）；Morlet连续小波变换结合1000次模拟的一阶自回归（AutoRegressive order 1, AR(1)）红噪声零假设；基于200次重复参数自助法的穷尽贝叶斯信息准则（Bayesian Information Criterion, BIC）断点搜索；以2005年韩国标准人口为基准的直接年龄标化；基于Queen邻域+KNN-2基线及11种备选权重方案的全局Moran's I检验（9999次置换检验）；带Benjamini-Hochberg错误发现率的局部空间关联指标（LISA）分析；经验贝叶斯（Marshall 1991）市郡区收缩率分析；分时段对比分析（2001-2012年与2013-2024年）。 【可复现性说明】 所有蒙特卡洛程序均采用种子值=42。运行环境为Python 3.12，依赖库包括pandas、numpy、statsmodels、scipy、geopandas、plotly、matplotlib、PyMannKendall。 【许可证】 知识共享署名4.0国际许可协议（CC-BY-4.0）