% Performs fast detrended fluctuation analysis on a nonstationary input signal to
% obtain an estimate for the scaling exponent.
%
% Useage:
% [alpha, intervals, flucts] = fastdfa(x)
% [alpha, intervals, flucts] = fastdfa(x, intervals)
% Inputs
%    x          - input signal: must be a row vector
% Optional inputs
%    intervals  - List of sample interval widths at each scale
%                 (If not specified, then a binary subdivision is constructed)
%
% Outputs:
%    alpha      - Estimated scaling exponent
%    intervals  - List of sample interval widths at each scale
%    flucts     - List of fluctuation amplitudes at each scale
%
% (c) 2006 Max Little. If you use this code, please cite:
% M. Little, P. McSharry, I. Moroz, S. Roberts (2006),
% Nonlinear, biophysically-informed speech pathology detection
% in Proceedings of ICASSP 2006, IEEE Publishers: Toulouse, France.

function [alpha, intervals, flucts] = fastdfa(x, varargin)

[xpts, ypts] = fastdfa_core(x, varargin{:});

% Sort the intervals, and produce a log-log straight line fit
datapts   = sortrows([xpts ypts],1);
intervals = datapts(:,1);
flucts    = datapts(:,2);

coeffs    = polyfit(log10(xpts), log10(ypts), 1);
alpha     = coeffs(1);