Visualization

Multivariate statistical methods are used to analyze the joint behavior of more than one random variable. There are a wide range of multivariate techniques available, as may be seen from the examples below. Statgraphics data visualization software functions as multivariate software that can perform these sorts of tasks.

Procedure Statgraphics Centurion Statgraphics
Sigma express
Statgraphics
stratus
Statgraphics
Web Services
StatBeans
Matrix Plot
Parallel Coordinates Plot        
Andrews Plot        
Star Glyphs and Sunray Plots        
Chernoff Faces        
matrixplot

Matrix Plot

Our data visualization software can produce matrix plots that are used to display all pairs of X-Y plots for a set of quantitative variables. With multivariate software, these are a good method for detecting pairs of variables that are strongly correlated. It is also possible to detect cases that appear to be outliers.
The matrix plot at the right has two additions:
1. A box-and-whisker plot for each variable in the diagonal locations.
2. A robust LOWESS smooth for each plot, which highlights the estimated relationships between the variables.

More:Matrix Plot.pdf

parallel

Parallel Coordinates Plot

The Parallel Coordinates Plot is a multivariate visualization technique that can be very useful in identifying differences and similarities amongst observed cases when the number of dimensions is too large to use a standard scatterplot using data visualization software. Statgraphics multivariate software uses the line for the i-th row to connect the standardized values of each variable in that row, where the standardized value equals the observed value minus the sample mean of that variable divided by the sample range.

More: Parallel Coordinates Plot.pdf

andrews

Andrews Plot

The Andrews Plot is similar to the Parallel Coordinates Plot, except that it encodes the information for each variable on a sine or cosine with different frequencies.

More:

glyphs

Star Glyphs and Sunray Plots

A glyph is a geometric object made with data visualization software that represents the values of each quantitative variable. The size of the polygon in each direction is scaled according to the value of a specific variable. Cases with similar characteristics will have similar shapes.

More:Star Glyphs and Sunray Plots.pdf

faces

Chernoff Faces

Chernoff Faces provide a method for visualizing multivariate data by drawing cartoon faces in which various features are scaled according to the values of different quantitative variables. They were developed by Herman Chernoff and first described in the article titled “The Use of Faces to Represent Points in k-Dimensional Space Graphically”, published in 1973. While their effectiveness as a method for identifying groups of cases has been debated, they represent a novel alternative to more conventional multivariate visualization techniques and can be made with Statgraphics multivariate software.

More: Chernoff Faces.pdf