We introduce a novel family of projected distributions on the circle and the sphere, namely the circular and spherical projected Cauchy distributions as promising alternatives for modeling circular and directional data. The circular distribution encompasses the wrapped Cauchy distribution as a special case, featuring a more convenient parameterisation. Next, we propose a generalised wrapped Cauchy distribution that includes an extra parameter, enhancing the fit of the distribution. In the spherical context, we impose two conditions on the scatter matrix, resulting in an elliptically symmetric distribution. Our projected distributions exhibit attractive properties, such as closed-form normalising constants and straightforward random value generation. The distribution parameters can be estimated using maximum likelihood and we assess their bias through numerical studies. We compare our proposed distributions to existing models with real data sets, demonstrating superior fit both with and without covariates.
翻译:我们引入了一种新颖的投影分布族,在圆和球面上,即循环和球形投影柯西分布,作为建模循环和方向数据的有前途的选择。圆分布包括缠绕柯西分布作为一种特殊情况,具有更方便的参数化。接下来,我们提出了一种通用的缠绕柯西分布,包括一个额外的参数,增强了分布的适合度。在球形情况下,我们对散布矩阵施加了两个条件,得到了一个椭圆对称分布。我们的投影分布具有有吸引力的特性,如闭合形式的常数和简单直接的随机值生成。可以使用最大似然估计分布参数,并通过数值研究评估其偏差。我们将我们提出的分布与现有模型在真实数据集上进行比较,证明在有或没有协变量的情况下都具有优越的拟合程度。