Radviz is a radial spring-based visualization that permits the visualization
of n-dimensional datasets. Data attributes are equidistantly distributed along
the circumference of a circle. Each data item is virtually
connected to a spring that starts at the circle perimeter and ends on the data item.
Each spring pulls the item with a force proportional to the item attribute value.
Depending on the value of each attribute, the forces of the springs project
each data item to a position inside the circle where the sum of the spring forces
is equal to zero.
Usage
radviz2d(dataset, name = "")
Arguments
dataset
The dataset to be visualized.
name
The name of the dataset to be used in the graph title.
Value
A Radviz visualization of the original dataset is returned.
Details
Some features of this visualization are:
l) Points where all dimensional values have approximately the same value will lie
close to the center.
2) If dimensional points lie opposite each other on the circle and have similar
values than points will lie near the center.
3) If 1 or 2 dimensional values are greater, points will lie closer to those
dimensional points.
4) Where a point will lie depends on the layout of the particular dimensions around
the circle.
5) This is a non-linear projection from N-dimensions down to 2 dimensions
6) Certain symmetries of the data will be preserved.
The function assumes the class labels are in the last column. Class column may be
either a numeric vector or a factor.
References
Ankerst M., Keim D. A., Kriegel H.-P. Circle Segments: A Technique for Visually
Exploring Large Multidimensional Data Sets, IEEE Visualization, 1996.
K.A. Olsen, R.R. Korfhage, K.M. Sochats, M.B. Spring and J.G. Williams. Visualisation of a
Document Collection: The VIBE System, Information Processing and Management, Vol. 29, No. 1, pp.
69-81, Pergamon Press Ltd, 1993.