NCL Website header
NCL Home > Documentation > Functions > Interpolation

idsfft

Interpolates 2D random data.

Prototype

	function idsfft (
		xc  [*] : float,    
		yc  [*] : float,    
		zc  [*] : float,    
		dim [2] : integer   
	)

	return_val [dim[0]][dim[1]] :  float

Arguments

xc

A one-dimensional array containing the x-coordinates of the input data. xc must have size four or greater.

yc

A one-dimensional array containing the y-coordinates of the input data. yc must have size four or greater and be the same size as xc.

zc

A one-dimensional array containing values at coordinates (xc,yc). zc must have size four or greater and be the same size as xc and yc.

dim

A one-dimensional array of size 2 containing the dimensions for the returned array, i.e. the returned array will be two-dimensional and have dimension sizes dim[0] and dim[1].

Return value

idsfft returns a float array of size dim[0] x dim[1]. The values in the returned array are interpolated from the input data. The x-coordinates for the returned array are determined by finding the span of the min and max values of the xc array and dividing that span into dim[0] equally-spaced coordinates. Similarly for the y-coordinates. Run the example below for clarification if desired.

Description

idsfft interpolates the data values zc, defined at coordinates (xc,yc), onto a regular grid.

idsfft returns a two-dimensional float array of size dim[0] x dim[1]. These values are interpolated from the input. The x-coordinates for the returned array are determined by finding the span of the min and the max of the xc array and dividing that span into dim[0] equally-spaced coordinates. Similarly for the y-coordinates. Run the example below for clarification if desired. The derived coordinate variables are attributes of the returned array. Dimension 0 of the output is named "ncl0" and dimension 1 of the output is named "ncl1".

Note: a serious bug was fixed in this routine in version 4.2.0.a31 (released mid-January 2004) in which incorrect results would be calculated for cases where dim[0] was not equal to dim[1].

There are many interpolation functions superior to idsfft. It is recommended you use one of those. For a list of other interpolation functions, see the list of interpolation routines.

Note: this function no longer exits when an error occurs.

See Also

List of interpolation routines

Examples

Example 1

Interpolation on random data.

begin
  xran = (/                                                   \
           12., 60., 14., 33.,  8., 12., 43., 57., 22., 15.,  \
           19., 12., 64., 19., 15., 55., 31., 32., 33., 29.,  \
           18.,  2., 18., 42., 56.,  9.,  6., 12., 44., 19.   \
         /)
  yran = (/                                                   \
            0.,  2.,  3., 53.,  7., 11., 13., 17., 19., 49.,  \
            1., 31., 37.,  5.,  7., 47., 61., 17.,  5., 23.,  \
            29.,  3.,  5., 41., 43.,  9., 13., 59.,  1., 67.  \
         /)
  zran = (/                                                   \
           1.0, 1.5, 1.7, 1.4, 1.9, 1.0, 1.5, 1.2, 1.8, 1.4,  \
           1.8, 1.7, 1.9, 1.5, 1.2, 1.1, 1.3, 1.7, 1.2, 1.6,  \
           1.9, 1.0, 1.6, 1.3, 1.4, 1.8, 1.7, 1.5, 1.1, 1.0   \
         /)

  out = (/50,60/)

  interp = idsfft(xran,yran,zran,out)

  print(interp&ncl0)  ;  50 equally-spaced coordinates in the range 2. to 64.
  print(interp&ncl1)  ;  60 equally-spaced coordinates in the range 0. to 67.
  print(interp)       ;  A 50 x 60 array of interpolated values.

end

Errors

If two coordinates (xc[i],yc[i]) and (xc[j],yc[j]) are equal for i not equal to j, an error is reported.

If any three coordinates in the input arrays (xc,yc) are collinear, an error is reported.