>  Docs Center  >  Libraries  >  ASTROLIB  >  CSPLINE
Libraries

CSPLINE

CSPLINE

Name


      CSPLINE

Purpose


      Function to evaluate a natural cubic spline at specified data points

Explanation


      Combines the Numerical Recipes functions SPL_INIT and SPL_INTERP

Calling Sequence


      result = cspline( x, y, t, [ DERIV = ])

Inputs


      x - vector of spline node positions, must be monotonic increasing or
          decreasing
      y - vector of node values
      t - x-positions at which to evaluate the spline, scalar or vector
  INPUT-OUTPUT KEYWORD:
      DERIV - values of the second derivatives of the interpolating function
              at the node points. This is an intermediate step in the
              computation of the natural spline that requires only the X and
              Y vectors. If repeated interpolation is to be applied to
              the same (X,Y) pair, then some computation time can be saved
              by supplying the DERIV keyword on each call. On the first call
              DERIV will be computed and returned on output.

Output


      the values for positions t are returned as the function value
      If any of the input variables are double precision, then the output will
      also be double precision; otherwise the output is floating point.

Example


      The following uses the example vectors from the SPL_INTERP documentation
      IDL> x = (findgen(21)/20.0)*2.0*!PI ;X vector
      IDL> y = sin(x) ;Y vector
      IDL> t = (findgen(11)/11.0)*!PI ;Values at which to interpolate
      IDL> cgplot,x,y,psym=1 ;Plot original grid
      IDL> cgplot, /over, t,cspline(x,y,t),psym=2 ;Overplot interpolated values

Method


      The "Numerical Recipes" implementation of the natural cubic spline is
      used, by calling the intrinsic IDL functions SPL_INIT and SPL_INTERP.

History


      version 1 D. Lindler May, 1989
      version 2 W. Landsman April, 1997
      Rewrite using the intrinsic SPL_INIT & SPL_INTERP functions
      Converted to IDL V5.0 W. Landsman September 1997
      Work for monotonic decreasing X vector W. Landsman February 1999



© 2024 NV5 Geospatial Solutions, Inc. |  Legal
   Contact Us