A biplot can be loosely defined as a plot that displays both the observations (rows) and variables (columns) of multivariate data in a single plot. Consider a data set in which the observations are car models and the variables are ratings on measures such as miles per gallon, reliability, and so on. For such data a biplot like the following plots the cars as points and the measures as vectors showing how the cars are associated with the measures.

Biplots can be produced for the following analyses:

• Principal components analysis as performed in SAS/IML^® Studio. See the example titled "Reduce Dimensionality through Principal Component Analysis" in the Multivariate Analysis: Principal Component Analysis chapter of the SAS/IML Studio User's Guide.
• Multidimensional preference analysis. Use the MDPREF or PLOTS=MDPREF option in PROC PRINQUAL. See the example in the Getting Started section of the PRINQUAL documentation.
• Correspondence analysis. PROC CORRESP produces a biplot by default. See the Getting Started section of the CORRESP documentation.
• Preference mapping. PROC TRANSREG can be used to show additional variables in the space produced by a procedure like PRINQUAL. See the example titled "Preference Mapping of Automobile Data" in the TRANSREG documentation.