rtables -- A Framework For Creating Complex Structured Reporting Tables Via Multi-Level Faceted Computations

Abstract: Tables form a central component in both exploratory data analysis and formal reporting procedures across many industries. These tables are often complex in their conceptual structure and in the computations that generate their individual cell values. We introduce both a conceptual framework and a reference implementation for declaring, generating, rendering and modeling such tables. We place tables within the existing grammar of graphics paradigm for general statistical visualizations. Our open source rtables software implementation utilizes these connections to facilitate an intuitive way to declare complex table structure and construct those tables from data. In the course of this work, we relax several constraints present in the traditional grammar of graphics framing. Finally, rtables models instantiated tables as tree structures, which allows powerful, semantically meaningful and self-describing queries and manipulations of tables after creation. We showcase our framework in practice by creating complex, realistic example tables.