Skip to content

Which statistical methods should I use?

4 messages · Mustafa Tufan Turp, Sarah Goslee, Jose Manuel Molina

#
Hi,

We're happy to try to help, but we really need either a specific R
question or a clearer statement of problem. For instance, I have
no idea what a NAO index is, or what its statistical properties might
be. You don't describe what kind of temperature and precipitation data
you have, or give a clear idea of just what you are trying to test.

R has good facilities for CCA. This article by Gonzalez et al. in the
Journal of Statistical Software might be of use:
http://www.jstatsoft.org/v23/i12/
along with the associated R package.

If you have questions about applying those methods after you
read the paper and the package help and work through the
examples, then please explain more clearly what you need so
that we can help you.

Sarah
On Fri, Mar 5, 2010 at 4:43 AM, Mustafa Tufan Turp <turcomania at gmail.com> wrote:

  
    
#
Mustafa,

I have been working with these kind of statistical relationships, and I 
find that using spatial Pearson cross-correlation is very helpful to 
analyze  hydroclimatological associations. You can see this abstract:
http://hydrologydays.colostate.edu/Abstracts_09/Molina_poster_abs.pdf
Additional to CCA packages in R, if you feel the correlation maps and 
information that I mention in that abstract can be useful for your work, 
I will be pleased to give you more details.
Jose.
Mustafa Tufan Turp wrote: