Statistical applications to second-rank tensors in magnetic fabric analysis
Statistical applications to second-rank tensors in magnetic fabric analysis
复制标题
磁性织物分析中二阶张量的统计应用
DOI：
10.1046/j.1365-246x.2000.00174.x
复制
发表时间：
2000
影响因子：
2.8
通讯作者：
W. Owens
中科院分区：
文献类型：
--
作者：
W. Owens
磁组构中的统计应用出现在测量和人口描述中。关于测量，Hext（1963）的理论在这里被应用于Constable和Tauxe（1990）发表的数据，因为数据被证明令人满意地符合Hext假设的正态分布阅读误差模型; Constable和Tauxe主张自助分析。Hext的理论允许评估测量方案的可重复性（可重复性描述了误差方差对织物轴相对于测量参考系的方向的独立程度）。另一种选择，更接近旋转，设计建议，说明这方面的重要性赫克特的工作。人口的描述可以是在磁组构的非标准化或标准化（平均磁化率）。规范化的描述几乎被普遍使用，但可能并不总是合适的：这里将区分形式化，
SUMMARY Statistical applications in magnetic fabrics arise in measurement and in population description. Regarding measurement, Hext’s (1963) theory is applied here to data published by Constable & Tauxe (1990), since the data are shown to fit satisfactorily to Hext’s assumed model of normally distributed reading errors; Constable & Tauxe advocated a bootstrap analysis. Hext’s theory allows the rotatability of the measuring scheme to be assessed (rotatability describes the degree of independence of error variance on the orientation of the fabric axes relative to the measurement reference frame). An alternative, more nearly rotatable, design is suggested, illustrating the importance of this aspect of Hext’s work. Population description may be in terms of magnetic fabrics that are unnormalized or normalized (by mean susceptibility). Normalized descriptions are used almost universally, but may not always be appropriate: the distinction is formalized here for