Re: The Principal/Principle Component Confusion


"Joel Daniels" <jdaniel4@xxxxxxxx> wrote in message
news:1172116311.975326.254470@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Feb 20, 9:22 pm, raipane...@xxxxxxxxx wrote:
On Feb 20, 5:31 pm, "Joel Daniels" <jdani...@xxxxxxxx> wrote:

A-parent-lee we cud use sum more general aide-you-caution.

Even UCL ("London's Global University") failed to write it rightly:

http://www.ucl.ac.uk/oncology/MicroCore/HTML_resource/PCA_1.htm

Cheers,
Joel Daniels

P.S. Recently my spill chuck utility felled too cache the ward
"chagrinned" (with two "n"s). Sins then, I have been moor care full,
and I try too not trust it two folly.

On Feb 18, 1:49 pm, "Greg Heath" <h...@xxxxxxxxxxxxxxxx> wrote:

The use of the incorrect term "Principle" when referring to
Principal Component Analysis still has not been squelched
by the CTP (Correct Terminology Police)!

A recent search in my 6 favorite newsgroups using the keywords
"principle-component" have revealed the following number of infected
threads:

20 sci.stat.consult
18 comp.ai.neural-nets
17 sci.stat.math
16 comp.soft-sys.matlab
7 sci.stat.edu
2 sci.math.num-analysis

Imagine what one would find if the search was
expanded to all newgroups or (Heaven forbid!)
... the Web!

Greg- Hide quoted text -

- Show quoted text -

In this paper they use Principle and Principal
interchangeably:http://ieeexplore.ieee.org/iel5/8662/27463/01222584.pdf

Please correct me if I am missing something, but grammatically I do
not believe that "principle component[s] analysis" makes much sense.
According to the OED "principle" is a noun not an adjective, so
"principle components" refers to the components of the principle,
rather than those components which are the principal ones.

To be sure, "principle components analysis" is becoming increasingly
common, as a Google Scholar search quickly shows, but "principal
components analysis" is linguistically correct.

Blessings,
Joel Daniels

P.S. My post above was not meant to make fun of the authors of the
2,000+ scholarly papers listed on Google that use "principle
components analysis." I was just trying to have fun with homophones
and near-homophones.


A lecturer at Yale has a webpage with Principle

http://www.yale.edu/ceo/Documentation/PCA_Outline.pdf

which gives a link to "A Tutorial on Principle Component Analysis by
Lindsay I. Smith (Cornell University)

http://csnet.otago.ac.nz/cosc453/student_tutorials/principal_components.pdf

despite the fact that the name of the webpage contains the word "principal".

There are numerous such examples (60,000+) containing both spellings eg

"Principle Component Analysis

Performs a principal components analysis on the given data matrix and
returns the results as an object of class `prcomp'."

http://astrostatistics.psu.edu/vostat/help/prcomp.html

I suspect that mathematicians are often too much in a hurry - and their
thoughts are rushing ahead of their pen/keyboard.

Nick


.