在貝氏統計分析上,WinBUGS軟體是最好的。它不僅僅操作簡易,WinBUGS網站上都有清楚的步驟,還有詳細的文件、幫助、指導、範例。在處理資料方面,它採用了MCMC 法裡的Gibbs抽樣和M-H演算的方法,在處理模型上,又有極大的靈活性,可處理困難的模型。安裝方面也相當的簡單,而且它是免費的,對於初學者來說,不僅可以很快的上手,它還可以配合其他軟體進行使用,許多軟體,如R、SAS、Excel、S-PLUS、Matlab等,這些都是處理統計上不可或缺的軟體,現在可以結合這些軟體一起使用,操作起來會更加得心應手。 而在本研究中,為了討論韋伯分配設限資料之貝氏分析,需要使用R軟體來操作,所以我們結合了R與WinBUGS一起使用,利用R軟體裡的R2WinBUGS套件,安裝後即可在R軟體中使用WinBUGS。最後,我們使用R來模擬資料並與WinBUGS一起去對韋伯分配設限資料的模型參數做分析研究。
WinBUGS is the best software to do the Bayesian analysis. It’s website has clear steps, and the software is free to use. It is not only easy to use, but it also has detailed documentation of data, help, guidance, and examples. In processing data, WinBUGS omits the computational complexity when it applies Markov chain Monte Carlo method to execute Gibbs Sample and Metropolis-Hastings Algorithm. In processing model, it also has great of flexibility to do difficulties model, and it is friendly to beginners. It can be run even from other mathematical and statistical software packages such as R, Matlab and Excel., S-PLUS, and Matlab. In this study, we generate a random sample of type I interval censored data using R, and then do analysis by WinBUGS via the R package R2WinBUGS, which is available on the CRAN site. Finally, we and WinBUGS to analysis type I interval censored data from Weibull distribution.