Togaware DATA MINING
Desktop Survival Guide
by Graham Williams
Google

Recenter

A common normalisation is to recenter and rescale our data. The simplest approach to do this is to subtract the mean value of a variable from each observation's value of the variable (to recenter the variable) and to then divide the values by the root-mean-square of the variable values ( $\sqrt{\sum{x_i^2}}/{n-1}$), which re-scales the variable back to a range within a few integer values around zero.

Rattle relies on the scale function from the base package to perform the re-centering:



> weather$RRC_Evaporation <- scale(weather$Evaporation)[,1]
> summary(weather$RRC_Evapration)



Length  Class   Mode 
     0   NULL   NULL

Note that the resulting mean is not precisely zero, but pretty close.



Copyright © 2004-2010 Togaware Pty Ltd
Support further development through the purchase of the PDF version of the book.
The PDF version is a formatted comprehensive draft book (with over 800 pages).
Brought to you by Togaware. This page generated: Sunday, 22 August 2010