view scripts/general/interpft.m @ 18754:0ede4dbb37f1

Overhaul interp1, interp2, interp3 functions. * NEWS: Announce change in 'cubic' interpolation method for interp2 to match Matlab. * bicubic.m: Use interp2 (..., "spline") in %!tests. * interp1.m: Improve docstring. Use switch statement instead of if/elseif tree for simpler code. Use more informative error message than 'table too short'. Add titles to demo plots. Add new demo block showing difference between 'pchip' and 'spline' methods. * interp2.m: Rewrite docstring. Use variable 'extrap' instead of 'extrapval' to match documentation. Use clearer messages in error() calls. Make 'cubic' use the same algorithm as 'pchip' for Matlab compatibility. Use Octave coding conventions regarding spaces between variable and parenthesis. Added input validation tests. * interp3.m: Rewrite docstring. Use clearer messages in error() calls. Make 'cubic' use the same algorithm as 'pchip' for Matlab compatibility. Simplify input processing. Rewrite some %!tests for clarity. Added input validation tests.
author Rik <rik@octave.org>
date Sun, 30 Mar 2014 14:18:43 -0700
parents d63878346099
children 4197fc428c7d
line wrap: on
line source

## Copyright (C) 2001-2013 Paul Kienzle
##
## This file is part of Octave.
##
## Octave is free software; you can redistribute it and/or modify it
## under the terms of the GNU General Public License as published by
## the Free Software Foundation; either version 3 of the License, or (at
## your option) any later version.
##
## Octave is distributed in the hope that it will be useful, but
## WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
## General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with Octave; see the file COPYING.  If not, see
## <http://www.gnu.org/licenses/>.

## -*- texinfo -*-
## @deftypefn  {Function File} {} interpft (@var{x}, @var{n})
## @deftypefnx {Function File} {} interpft (@var{x}, @var{n}, @var{dim})
##
## Fourier interpolation.  If @var{x} is a vector, then @var{x} is
## resampled with @var{n} points.  The data in @var{x} is assumed to be
## equispaced.  If @var{x} is a matrix or an N-dimensional array, the
## interpolation is performed on each column of @var{x}.  If @var{dim} is
## specified, then interpolate along the dimension @var{dim}.
##
## @code{interpft} assumes that the interpolated function is periodic,
## and so assumptions are made about the endpoints of the interpolation.
##
## @seealso{interp1}
## @end deftypefn

## Author: Paul Kienzle
## 2001-02-11
##    * initial version
## 2002-03-17 aadler
##    * added code to work on matrices as well
## 2006-05-25 dbateman
##    * Make it matlab compatiable, cutting out the 2-D interpolation

function z = interpft (x, n, dim)

  if (nargin < 2 || nargin > 3)
    print_usage ();
  endif

  if (! (isscalar (n) && n == fix (n)))
    error ("interpft: N must be a scalar integer");
  endif

  if (nargin == 2)
    if (isrow (x))
      dim = 2;
    else
      dim = 1;
    endif
  endif

  nd = ndims (x);

  if (dim < 1 || dim > nd)
    error ("interpft: invalid dimension DIM");
  endif

  perm = [dim:nd, 1:(dim-1)];
  x = permute (x, perm);
  m = rows (x);

  inc = ceil (m/n);
  y = fft (x) / m;
  k = ceil (m / 2);
  sz = size (x);
  sz(1) = n * inc - m;

  idx = repmat ({':'}, nd, 1);
  idx{1} = 1:k;
  z = cat (1, y(idx{:}), zeros (sz));
  idx{1} = k+1:m;
  z = cat (1, z, y(idx{:}));

  ## When m is an even number of rows, the FFT has a single Nyquist bin.
  ## If zero-padded above, distribute the value of the Nyquist bin evenly
  ## between the new corresponding positive and negative frequency bins.
  if (sz(1) > 0 && k == m/2)
    idx{1} = n * inc - k + 1;
    tmp = z(idx{:}) / 2;
    z(idx{:}) = tmp;
    idx{1} = k + 1;
    z(idx{:}) = tmp;
  endif

  z = n * ifft (z);

  if (inc != 1)
    sz(1) = n;
    z = inc * reshape (z(1:inc:end), sz);
  endif

  z = ipermute (z, perm);

endfunction


%!demo
%! clf;
%! t = 0 : 0.3 : pi;  dt = t(2)-t(1);
%! n = length (t);  k = 100;
%! ti = t(1) + [0 : k-1]*dt*n/k;
%! y = sin (4*t + 0.3) .* cos (3*t - 0.1);
%! yp = sin (4*ti + 0.3) .* cos (3*ti - 0.1);
%! plot (ti, yp, 'g', ti, interp1(t, y, ti, "spline"), 'b', ...
%!       ti, interpft (y, k), 'c', t, y, "r+");
%! legend ("sin(4t+0.3)cos(3t-0.1)", "spline", "interpft", "data");

%!shared n,y
%! x = [0:10]';  y = sin(x);  n = length (x);
%!assert (interpft (y, n), y, 20*eps);
%!assert (interpft (y', n), y', 20*eps);
%!assert (interpft ([y,y],n), [y,y], 20*eps);

%% Test case with complex input from bug #39566
%!test
%! x = (1 + j) * [1:4]';
%! y = ifft ([15 + 15*j; -6; -1.5 - 1.5*j; 0; -1.5 - 1.5*j; -6*j]);
%! assert (interpft (x, 6), y, 10*eps);

%% Test for correct spectral symmetry with even/odd lengths
%!assert (max (abs (imag (interpft ([1:8], 20)))), 0, 20*eps);
%!assert (max (abs (imag (interpft ([1:8], 21)))), 0, 21*eps);
%!assert (max (abs (imag (interpft ([1:9], 20)))), 0, 20*eps);
%!assert (max (abs (imag (interpft ([1:9], 21)))), 0, 21*eps);

%% Test input validation
%!error interpft ()
%!error interpft (1)
%!error interpft (1,2,3)
%!error <N must be a scalar integer> interpft (1,[2,2])
%!error <N must be a scalar integer> interpft (1,2.1)
%!error <invalid dimension DIM> interpft (1,2,0)
%!error <invalid dimension DIM> interpft (1,2,3)