一本书：survival analysis生存分析

3. Prentice, Williams and Peterson (PWP) model
   (1) total time
   (2) gap time

我一直在理解这两种模型之间的异同。不知哪位有此答案？

以上模型在估计第一个事件时的参数应无差异。缘何我的模型就有差异？恳请指教

Many distinct evalations exist about  any priority on one model over others and no agreement has been reached so far, which forces users in an undecided situation how to select the best (or better) models in rcurrent analysis.
Anyone has opinion on that?

偶也上当了，这本书其实偶早就有了。

24# biostat

So sorry to hear about that. I know exactly the feeling when I was pushed or misguilded to pay for something useless.
On the other hand, the book is invaluable when one needs it and that is reason why I bought it.
Whatever, what can I do for you then?

回帖是是种美德。。。。。。。。。。。。。。

好书要支持，不过论坛币不过好头痛

Apparently, survival analyses are not hot here, a place being dominated by bank related modeling and programming. There it is so 'pure' on solely  being based on numbers where it is too complicated to go nowhere. In person, I dislike that with no specific reason even I know so many bright people are dedicated themselve on there.

Any different idea?

Recurrent events analyses in PROC PHREG can also work on time-varying variables with the form of counting process data input. That was confrimed by the SAS stats support and an expert in multiple-event and informative-censoring survival analysis. But how come some precision was lost in the analysis for the 1st event?

好像没有很多人对生存分析感兴趣啊。自己做了东西也不知道对不对，还是理论跟不上，只能照猫画虎。