Re: ncl-talk Digest, Vol 10, Issue 11

From: alexg (alexg AT maths.unsw.edu.au)
Date: Tue Apr 19 2005 - 18:44:05 MDT


Hi,
I'm trying to calculate a (95%) confidence level on a
correlation/regression map (ie correlating a single timeseries with the
timeseries for another variable, over a number of spatial points).
To do this for two timeseries I would use information about the
autocorrelations in order to work out an effective number of degrees of
freedom. Hoverver for a correlation map this is quite a computationally
intensive process - looping over each grid point and at multiple lags.
Are there any in built functions that do this?
Many thanks
Alex

_________________________________________________________________
   Alex Sen Gupta
   Centre for Environmental Modelling and Prediction (CEMAP)
   School of Mathematics
   The University of New South Wales Telephone: +61-2-9385-6902
   UNSW SYDNEY NSW 2052 Facsimile: +61-2-9385-7123
   Australia E-mail: alexg@maths.unsw.edu.au
   www.maths.unsw.edu.au/~alexg/
 _________________________________________________________________

_______________________________________________
ncl-talk mailing list
ncl-talk@ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk



This archive was generated by hypermail 2b29 : Tue Apr 19 2005 - 21:18:00 MDT