IPTA DR2 - GWB analysis MCMC output

IPTA DR2 common red noise, MCMC output

These files are the primary output from a Markov chain Monte Carlo (MCMC) sampling process. They are samples from the posterior probability distribution for a particular model described in the companion paper.

Each zipped tarball contains four files.  The "chain" file has several tab-separated columns, each of which corresponds to a model parameter, except the last four which are metadata.  The parameter names (including metadata) are listed in the companion "params" file.  The frequencies used in the common red noise models are listed in the "crn_frequencies" file.  Additional information is provided in a README file.

Each row of the chain file is one sample from the model posterior.  The first samples at the beginning of the MCMC are the "burn-in" phase, before the chain has converged to the posterior.  We recommend discarding the first ~25% of samples before using them to make inferences.

IPTA DR2 公共红噪声（common red noise）马尔可夫链蒙特卡洛（MCMC）输出结果

本文件集为马尔可夫链蒙特卡洛（Markov Chain Monte Carlo, MCMC）采样流程的核心输出产物，其样本源自配套论文中所描述的特定模型的后验概率分布。

每个压缩tar包均包含四个文件。其中「chain」文件包含若干制表符分隔的列，除最后四列为元数据外，其余每一列均对应一个模型参数。模型参数（含元数据）的名称详见配套的「params」文件；公共红噪声模型所使用的频率列表收录于「crn_frequencies」文件；额外说明信息则存放在README文件中。

chain文件中的每一行均代表模型后验分布的一组采样样本。MCMC流程初始阶段的前若干样本属于「预烧（burn-in）」阶段，此时马尔可夫链尚未收敛至后验分布。我们建议在利用样本开展推断分析前，先丢弃前约25%的采样数据。