Conferences CIMPA, 8th Latin American Conference on Statistical Computing (LACSC)

Font Size: 
New tools for visualizing distributional datasets
Corrado Crocetta, Antonio Irpino

Last modified: 2024-05-14

Abstract


Visualization tools are crucial for conveying patterns and guiding analytical approaches in data interpretation. When faced with the complexity of visualizing distributional data tables, where each observation is a vector of frequency or density distributions, the need for user-friendly tools becomes apparent. To address this challenge, three innovative visualization tools have been introduced for data tables characterized by numeric symbolic distributional data.The first two tools, the Green Eye Iris (GEI) and the Flower plot, utilize a polar coordinate-based representation of stacked bar charts or violin plots. The third tool extends the traditional heatmap plot and is particularly effective for illustrating datasets with numerous observations and variables. All three methods focus on visually representing the proportion of mass distributed on the domain variable, utilizing diverging color palettes for each distribution.

Keywords


Symbolic data, distributional data, visualization