For background have a look at http://en.wikipedia.org/wiki/Multicollinearity.
I have also used
Regression Diagnostics: Identifying Influential Data and Sources of
Collinearity (Wiley Series in Probability and Statistics) by David A.
Belsley, Edwin Kuh and Roy E. Welsch
Sections 1.9 to 1.12 of
Hands-On Intermediate Econometrics Using R: Templates for Extending
Dozens of Practical Examples [With CDROM] by Hrishikesh D. Vinod
(2008)
Basically how you proceed depends a lot on what you are trying to achieve.
Best Regards John
On 5 August 2012 23:04, Roberto Moscetti <rmoscetti at unitus.it> wrote:
Hi,
thank you for your help. I know, I need to learn enough statistics to
understand how to process my data. The reason because of I write on this
forum is to ask to people a way to learn.
I am a postharvest researcher and statistic is not my main field, so I try
to do my best.
Do you know a book (or literature) than can help me?
Thank you very much for your time and suggestions.
Best regards,
Roberto
Il 05/08/2012 12:55, Jeff Newmiller ha scritto:
There is no "magic bullet" (package) for your problem. You must either
learn enough statistics to understand how to analyze your data, or consult
with someone who does.
FWIW collinearity is not in general amenable to automatic removal.
However, you can identify which inputs are collinear with each other, and
omit the redundant ones next iteration of your analysis, using (for example)
the approach suggested by Uwe. Deciding WHICH of the redundant inputs is
most appropriate to keep is the part computers are not so good at... that is
where you must be smarter or more creative than the computer.
Also, it would help you get responses if you included the context (earlier
discussion) in your replies.. most people do not use Nabble here. Reading
and following the requests in the footer of every message will also help.
---------------------------------------------------------------------------
Jeff Newmiller The ..... ..... Go
Live...
DCN:<jdnewmil at dcn.davis.ca.us> Basics: ##.#. ##.#. Live
Go...
Live: OO#.. Dead: OO#.. Playing
Research Engineer (Solar/Batteries O.O#. #.O#. with
/Software/Embedded Controllers) .OO#. .OO#.
rocks...1k
---------------------------------------------------------------------------
Sent from my phone. Please excuse my brevity.
Roberto <rmoscetti at unitus.it> wrote:
I do not know, because I tried to use rfe function (Backwards Feature
Selection, Caret Package) to select wavelengths useful for a prediction
model. Otherwise, rfe function give me back a lot of warning messages
about
collinearity between variables.
So, I do not know if your script can be useful.
I tried to use VIF-Regression to select variables, but rfe function
advise
me with the same warning messages again.
What do you think about that?
Thank you very much for your help.
Best,
Roberto
--
View this message in context:
http://r.789695.n4.nabble.com/Package-to-remove-collinear-variables-tp4639200p4639226.html
Sent from the R help mailing list archive at Nabble.com.