Cached PyMC inference results for SOFC-NSR hierarchical Bayesian competing-risk analysis

This dataset contains the main ASR (area-specific resistance) model posterior inference results (idata_asr_v3.nc) for the paper: Park, EunJoo; Kwon, Hyochan; Lee, Jinkwang (2026). "Start–Stop Cycle-Induced Failure-Mode Transition in SOFC-Powered Northern Sea Route Shipping: A Hierarchical Bayesian Competing-Risk Analysis." Journal of Marine Science and Engineering (under review). The file is an ArviZ InferenceData object stored in NetCDF4 format, containing 4 chains × 1500 post-warmup draws of the hierarchical dual-degradation ASR model. It accompanies the source code repository at: https://github.com/parkej1130-dev/sofc-nsr-bayesian-competing-risk This file is hosted on Zenodo separately because its size (26 MB) exceeds the GitHub web upload limit. To use it, place it in the `results/` directory of the GitHub repository and run `verify_results.py` to reproduce the published numerical values (Tables 2, 3, and Section 3.6 of the main paper). Software stack: PyMC 5.28.4, ArviZ 0.22.0, NumPy 2.0.2, SciPy 1.16.3, Python 3.12.13.本数据集包含对应论文的核心专属区域阻力（area-specific resistance, ASR）模型后验推断结果文件idata_asr_v3.nc。对应论文信息如下：Park, EunJoo、Kwon Hyochan、Lee Jinkwang（2026），《启停循环诱导的固体氧化物燃料电池（Solid Oxide Fuel Cell, SOFC）驱动北极航线船舶失效模式转变：分层贝叶斯竞争风险分析》，《海洋科学与工程学报》（"Journal of Marine Science and Engineering"，审稿中）。 该文件为采用NetCDF4格式存储的ArviZ推理数据（ArviZ InferenceData）对象，包含分层双降解ASR模型的4条马尔可夫链×1500次预热后采样结果。 本数据集配套的源代码仓库地址为：https://github.com/parkej1130-dev/sofc-nsr-bayesian-competing-risk。 由于该文件大小为26 MB，超出GitHub网页端上传限制，因此单独托管于Zenodo平台。使用时，请将该文件放置至该GitHub仓库的`results/`目录下，并运行`verify_results.py`以复现论文中已发表的数值结果（对应论文表2、表3及第3.6节内容）。 所用软件栈如下：PyMC 5.28.4、ArviZ 0.22.0、NumPy 2.0.2、SciPy 1.16.3及Python 3.12.13。