Discovering Fuzzy and Statistical Patterns in Data: The nuggets R Package

The nuggets package provides a flexible and extensible frame-work for discovering interpretable data patterns based on frequent logical conditions. Its designunifies classical association-rule mining with linguistic and fuzzy representations, while enablingoptional statistical evaluation for selected pattern types such as conditional contrasts and corre-lations. Pattern generation is driven by support, ensuring efficient mining of relevant conditions,whereas additional quantitative analyses or tests can be seamlessly attached when desired.A major strength of nuggets lies in its extensibility. The framework allows users to definecustom fuzzification schemes and to evaluate an arbitrary R function on every frequent con-dition, thereby enabling the creation of new, user-defined pattern types. This design encour-ages experimentation with alternative logical semantics, statistical measures, and application-specific evaluation criteria, making nuggets not only a tool for applied pattern discovery butalso a research platform for developing new methods.