#include "cctk.h" SUBROUTINE intp3d ( x, y, z, f, kt, ft, nx, ny, nz, xt, yt, zt, . d1, d2, d3 ) c implicit none c c--------------------------------------------------------------------- c c purpose: interpolation of a function of three variables in an c equidistant(!!!) table. c c method: 8-point Lagrange linear interpolation formula c c x input vector of first variable c y input vector of second variable c z input vector of third variable c c f output vector of interpolated function values c c kt vector length of input and output vectors c c ft 3d array of tabulated function values c nx x-dimension of table c ny y-dimension of table c nz z-dimension of table c xt vector of x-coordinates of table c yt vector of y-coordinates of table c zt vector of z-coordinates of table c c d1 centered derivative of ft with respect to x c d2 centered derivative of ft with respect to y c d3 centered derivative of ft with respect to z c Note that d? only make sense when intp3d is called with kt=1 c--------------------------------------------------------------------- c c c integer kt,nx,ny,nz,ktx double precision x(kt),y(kt),z(kt),f(kt) double precision xt(nx),yt(ny),zt(nz) double precision ft(nx,ny,nz) double precision d1,d2,d3 c c PARAMETER (ktx = 400) double precision fh(ktx,8), delx(ktx), dely(ktx), delz(ktx), & a1(ktx), a2(ktx), a3(ktx), a4(ktx), & a5(ktx), a6(ktx), a7(ktx), a8(ktx) double precision dx,dy,dz,dxi,dyi,dzi,dxyi,dxzi,dyzi,dxyzi integer n,ix,iy,iz IF (kt .GT. ktx) call CCTK_WARN (0, '***KTX**') c c c------ determine spacing parameters of (equidistant!!!) table c dx = (xt(nx) - xt(1)) / FLOAT(nx-1) dy = (yt(ny) - yt(1)) / FLOAT(ny-1) dz = (zt(nz) - zt(1)) / FLOAT(nz-1) c dxi = 1. / dx dyi = 1. / dy dzi = 1. / dz c dxyi = dxi * dyi dxzi = dxi * dzi dyzi = dyi * dzi c dxyzi = dxi * dyi * dzi c c c------- loop over all points to be interpolated c DO n = 1, kt c c------- determine location in (equidistant!!!) table c ix = 2 + INT( (x(n) - xt(1) - 1.e-10) * dxi ) iy = 2 + INT( (y(n) - yt(1) - 1.e-10) * dyi ) iz = 2 + INT( (z(n) - zt(1) - 1.e-10) * dzi ) c ix = MAX( 2, MIN( ix, nx ) ) iy = MAX( 2, MIN( iy, ny ) ) iz = MAX( 2, MIN( iz, nz ) ) c c write(*,*) iy-1,iy,iy+1 c c------- set-up auxiliary arrays for Lagrange interpolation c delx(n) = xt(ix) - x(n) dely(n) = yt(iy) - y(n) delz(n) = zt(iz) - z(n) c fh(n,1) = ft(ix , iy , iz ) fh(n,2) = ft(ix-1, iy , iz ) fh(n,3) = ft(ix , iy-1, iz ) fh(n,4) = ft(ix , iy , iz-1) fh(n,5) = ft(ix-1, iy-1, iz ) fh(n,6) = ft(ix-1, iy , iz-1) fh(n,7) = ft(ix , iy-1, iz-1) fh(n,8) = ft(ix-1, iy-1, iz-1) c c------ set up coefficients of the interpolation polynomial and c evaluate function values c a1(n) = fh(n,1) a2(n) = dxi * ( fh(n,2) - fh(n,1) ) a3(n) = dyi * ( fh(n,3) - fh(n,1) ) a4(n) = dzi * ( fh(n,4) - fh(n,1) ) a5(n) = dxyi * ( fh(n,5) - fh(n,2) - fh(n,3) + fh(n,1) ) a6(n) = dxzi * ( fh(n,6) - fh(n,2) - fh(n,4) + fh(n,1) ) a7(n) = dyzi * ( fh(n,7) - fh(n,3) - fh(n,4) + fh(n,1) ) a8(n) = dxyzi * ( fh(n,8) - fh(n,1) + fh(n,2) + fh(n,3) + & fh(n,4) - fh(n,5) - fh(n,6) - fh(n,7) ) c d1 = -a2(n) d2 = -a3(n) d3 = -a4(n) f(n) = a1(n) + a2(n) * delx(n) & + a3(n) * dely(n) & + a4(n) * delz(n) & + a5(n) * delx(n) * dely(n) & + a6(n) * delx(n) * delz(n) & + a7(n) * dely(n) * delz(n) & + a8(n) * delx(n) * dely(n) * delz(n) c ENDDO c RETURN END