目的在临床随访研究中,基于时间尺度指标的限制平均生存时间(restricted mean survival time,RMST)越来越受到关注,然而目前基于RMST的统计推断主要用于两组比较,缺少进行两组以上比较的方法。方法本文提出RMST多组间的假设检验法,包括...目的在临床随访研究中,基于时间尺度指标的限制平均生存时间(restricted mean survival time,RMST)越来越受到关注,然而目前基于RMST的统计推断主要用于两组比较,缺少进行两组以上比较的方法。方法本文提出RMST多组间的假设检验法,包括经典法(naive)、对数转换法(log)、双对数转换法(cloglog)三种检验法,并通过Monte Carlo模拟评价其Ⅰ类错误和检验效能,最后进行实例分析。结果综合Monte Carlo模拟的Ⅰ类错误及检验效能结果,显示所提出的RMST检验可以处理多组比较的问题,特别是cloglog转换法最为稳健。结论针对生存数据的多组比较问题,若考虑从时间尺度指标分析,推荐使用cloglog转换法的RMST多组检验。展开更多
We analyze the phenomena of phase group synchronization between the different nominal frequency signals and propose a new theory of the equivalent comparison between them. The exact expression of the equivalent compar...We analyze the phenomena of phase group synchronization between the different nominal frequency signals and propose a new theory of the equivalent comparison between them. The exact expression of the equivalent comparison is deduced. High resolution frequency measurement and phase comparison can be realized using this theory with the divider. For avoiding the frequency mixing, multiplication and synthesis, the system phase noise is improved and the higher resolution comparison and measurement are achieved between the different nominal frequencies by theory.展开更多
基金supported by the National Natural Science Foundation of China(Grant Nos.10978017 and 61201288)the Fundamental Research Funds for the Central Universities,China(Grant No.JB140413)
文摘We analyze the phenomena of phase group synchronization between the different nominal frequency signals and propose a new theory of the equivalent comparison between them. The exact expression of the equivalent comparison is deduced. High resolution frequency measurement and phase comparison can be realized using this theory with the divider. For avoiding the frequency mixing, multiplication and synthesis, the system phase noise is improved and the higher resolution comparison and measurement are achieved between the different nominal frequencies by theory.