| implementing pre pritten routines [message #39853] |
Sat, 19 June 2004 14:34 |
aontman1
Messages: 2
Registered: June 2004
|
Junior Member |
|
|
Hi,
I was wondering if anyone could help me. I was given pre written
routine
this is what is given to me (by the name of the author i believe it
came off one the messages here)
I have 3 dimensional array of data (100x40x10)
I was wondering if someone could help me write a code that would
interpolate the data (I hope that makes some sense). Any other
suggestions are welcome
Thank you so much for taking your time to answer my silly question
Lafoz
; PURPOSE:
;
; Perform n-dimensional linear interpolation on arrays of
; arbitrary dimensionality (ala the bilinear and trilinear
; interpolation of INTERPOLATE).
;
; CALLING SEQUENCE:
;
; value=ninterpolate(data,point)
;
; INPUTS:
;
; DATA: The n-dimensional array to interpolate.
;
; POINT: A vector of length n specifying the (single) point for
; which the interpolated value is desired.
;
; OUTPUTS:
; The interpolated value at the specified point.
this is the code:
function ninterpolate,data,point
n=n_elements(point) & two_n=2L^n & reb=[n,two_n]
if n ne size(data,/N_DIMENSIONS) then $
message,'Point must specify 1 coordinate for each dimension'
if n eq 1 then return, interpolate(data,point[0])
inds=(rebin(indgen(1,two_n),reb) AND
rebin([ishft(1,indgen(n))],reb)) gt 0
base=floor(point) & f=float(point)-base
f=[(1.-f),f]
multi_to_single=[1,product((size(data,/DIMENSIONS))[0:n-2],/ CUMULATIVE)]
data_ind=rebin(base,reb)+inds
data_ind=long(total(data_ind*rebin(multi_to_single,reb),1))
return, total(product(f[rebin(indgen(n),reb)+inds*n],1)*data[data_in d])
end
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
