Radiatrix

A columnar multi-dimensional data visualization tool

View the Project on GitHub whorvath2/Radiatrix

Welcome to the Radiatrix

Overview

The Radiatrix allows tuples (measures with any arbitrary number of degrees of freedom) that are taken repeatedly (either across or within subjects) to be visualized using a combination of polar and axial perspectives, i.e. arranged as a column or cylinder.

For the purpose of visualization, the scale of units of individual measures don't matter, especially in relation to other measures into which they could be logically converted. The idea of this tool is to show how measures correlate with one another, regardless of the scale at which they are constructed. Ergo, all measures are scaled to be 'visually appreciable' in the viewing scene.

To clarify with an example: If measures were taken in one dimension in millimeters, and along another dimension in meters, both will be represented in the radiatrix as a ratio within their maximum measures, and not relative to one another's actual scale.

Specification

Each tuple is represented in a polar coordinate space. The circle is arbitrarily divided into equal-sized portions ("slices") by as many lines (or radii) are there are degrees of freedom in the tuple. Each line is marked according to its units, and serve as a scale along which vectors are drawn for each data point, with the vectors originating in the center of the circle, and ending at the value corresponding to that in the tuple's data point.

In any arbitrary experiment resulting in a set of tuples, each set's data is plotted on its own polar coordinate space. The spaces from all or groups of the tuples can then be "stacked" into a column for visualization purposes, subject to the following restrictions: