Hi,
SITUATION:
I have model data on a (40layer x 400lat x 550lon) grid, converted out
of a rotated lat/lon grid. I have lat2D and lon2D fields for this data.
I want to extract whole vertical columns of this data at a set of
specific coordinates (lat/lon). I use the function 'getind_latlon2d' to
give me the indices into lat2D/lon2D for the nearest grid points to
these locations, which is acceptable.
PROBLEM:
The output of 'getind_latlon2d' has the form (n_points x 2), a set of
indices into the rightmost two dimensions of my data. I want to use
these indices to subscript into the data to give me a set of columnar
data for each of the locations, i.e. a 2D array. However, when I
subscript the data in the form 'data( : , idx(:,0) , idx(:,1) ), I get a
3D array in return. What I actually want is the horizontal diagonal of
this array (:,0,0), (:,1,1,),(:,2,2) etc.
QUESTION:
Is there a sensible way of extracting a set of vertical columns out of
3D data and writing them into a 2D array using proper array indexing
notation? Yes of course this can be done with loops but that is horribly
slow and not really the point, is it?
Thanks for any tips,
Jakob
-- Jakob Tendel Phone 303 497 2713 Research Applications Laboratory Office 3106b National Center for Atmospheric Research Building FL2 3450 Mitchell Lane Boulder CO 80301
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2.1.8 : Wed Jun 16 2010 - 15:28:33 MDT