Menke, W. (2012), Geophysical Data Analysis: Discrete Inverse Theory: Chapter 10, Factor Analysis

A suite of igneous rock samples from the floor of the Atlantic Ocean, obtained from PetDB, has systematic variation in major element chemical composition, as shown by the scatter plots (Figure 1). However, while scatter plots of pairs of elements reveal intriguing patterns of variation, they are limited by being two dimensional, whereas the chemical variation is multidimensional. Consequently, all the plots look different, making a unified interpretation difficult. Factor analysis indicates that only three dimensions (C1, C2, C3) are necessary to capture the most significant chemical variability. When plotted on these axes (Figure 2), a much simpler pattern is evident.

-W. Menke

 

Menke, William. Geophysical Data Analysis: Discrete Inverse Theory (Third Edition), 2012

 

View the book on Elsevier ScienceDirect

 

 

Tags: