POLINT

     Interpolate a set of N points by fitting a polynomial of degree N-1

     Adapted from algorithm in Numerical Recipes, Press et al. (1992),
     Section 3.1.

     POLINT, xa, ya, x, y, [ dy ]

     XA - X Numeric vector, all values must be distinct.   The number of
          values in XA should rarely exceed 10 (i.e. a 9th order polynomial)
     YA - Y Numeric vector, same number of elements
     X - Numeric scalar specifying value to be interpolated

     Y - Scalar, interpolated value in (XA,YA) corresponding to X

     DY - Error estimate on Y, scalar

     Find sin(2.5) by polynomial interpolation on sin(indgen(10))
               IDL> xa = indgen(10)
               IDL> ya = sin( xa )
               IDL> polint, xa, ya, 2.5, y ,dy
     The above method gives  y = .5988 & dy = 3.1e-4  a close
     approximation to the actual sin(2.5) = .5985

     Uses Neville's algorithm to iteratively build up the correct
     polynomial, with each iteration containing one higher order.

     Written W. Landsman                 January, 1992
     Converted to IDL V5.0   W. Landsman   September 1997

[Home Page] [Software, Documentation] [IDL Documentation] [Quick Reference] [Feedback]