A0373
Title: Goodness-of-fit and clustering of spherical and directional data: A comprehensive R package
Authors: Giovanni Saraceno - University at Buffalo (United States) [presenting]
Marianthi Markatou - University at Buffalo (United States)
Abstract: A new R package is presented that encodes innovative methodologies for data analysis. The package offers a comprehensive implementation of goodness-of-fit tests and clustering techniques based on quadratic distances. One-sample tests and two-sample tests for assessing the fit of probability distributions are implemented. Furthermore, tests for uniformity on the d-dimensional sphere based on Poisson kernel densities, provide additional capabilities. The package incorporates a clustering algorithm designed for data that can be analyzed as spherical data. By leveraging a mixture of Poisson-kernel-based densities on the sphere, the method facilitates effective clustering of spherical (or spherically transformed) data, providing insights into the underlying patterns and relationships. In summary, the proposed R package encompasses a suite of tools through which researchers and practitioners can gain deeper insights, make robust inferences, and provide potentially impactful analyses across diverse fields.
