Skip to content
Back to formatted view

Raw Message

Message-ID: <20030614131007.GA26789@mail2.sas.upenn.edu>
Date: 2003-06-14T13:10:07Z
From: Jonathan Baron
Subject: Normalizing data
In-Reply-To: <B6E43BA0-9E67-11D7-B0BC-0005026E2B43@vanderkogel.net>

On 06/14/03 14:57, Jonck van der Kogel wrote:
>Hello dear R-list members,
>I have a question about normalizing data. The goal is to normalize the 
>dataset per column, so all the data in each column is scaled to the 
>interval (0,1), will have a mean of 0 and a standard deviation of 1.

In psychology, we usually call this standardizing.  "Normalizing"
is subtracting the mean but NOT dividing by the s.d.

>I know the way to do this is to take each datapoint, subtract the mean 
>of the column it is located in and divide this by the standard 
>deviation of the column. Now my question is: is there a function in R 
>that does this, and if so, which function?

scale()

It will standardize (default) or normalize. And the default is to
it by column, as you describe.

-- 
Jonathan Baron, Professor of Psychology, University of Pennsylvania
R page:               http://finzi.psych.upenn.edu/