movingaverage.m

有监督自组织映射-偏最小二乘算法(A supervised self-organising map–partial least squares algorithm),可以用语多变量数据的回归分析

function [Data_out] = MovingAverage(Data_in, size_lag, size_lead, ma_mode)
% [Data_out] = MovingAverage(Data_in, size_lag, size_lead)
%
% Moving Average Filtering (row-wise)
% 
%
% INPUT:
% - Data_in              : The specified variable not to remove
% - size_lag (optional)  : Size of the lag (default is 2)
% - size_lead (optional) : Size of the lead (default is 2)
%
%
%
%  Author:   P.W.T. Krooshof & W.J. Melssen
%            Radboud University Nijmegen
%            Institute for Molecules and Materials
%            Analytical Chemistry
%            Nijmegen, The Netherlands
%  Created:  February 28, 2006
%  Updated:  

if nargin < 3
    size_lag = 2;
    size_lead = 2;
    ma_mode = 'mean';
end

num_points = size(Data_in, 2);

for j = 1:num_points
    switch lower(ma_mode)
        case 'mean'
            Data_out(:,j) = mean(Data_in(:,[(max(1,j-size_lag)):(min(num_points,j+size_lead))]),2);
        otherwise
            Data_out(:,j) = median(Data_in(:,[(max(1,j-size_lag)):(min(num_points,j+size_lead))]),2);
    end
end