samer@4: function [Y,M]=addnorm(F,X)
samer@4: % addnorm - additive normalisation with respect to arbitrary function
samer@4: % 
samer@4: % addnorm :: 
samer@4: %    ([[N,M]] -> [[N,1]]) | ([[N,M]] -> [[1,M]]) 
samer@4: %       ~'function to compute offsets',
samer@4: %    [[N,M]]           ~'data to normalise'
samer@4: % -> [[N,M]]           ~'offset vectors' 
samer@4: %    [[N,1]] | [[1,M]] ~'the vector that was subtracted'.
samer@4: %
samer@4: % This function can work row-wise OR column-wise depending on what
samer@4: % the offset computing function returns.
samer@4: %
samer@4: % EXAMPLES
samer@4: % 
samer@4: % To zero-mean each ROW of an N-by-M array X, use
samer@4: %    addnorm(@(t)mean(t,2),X);
samer@4: %
samer@4: % To zero the MEDIAN of each COLUMN, use
samer@4: %    addnorm(@median,X);
samer@4: 
samer@4: Y=vecop1(@minus,X,F(X)); % vecshift(F(X),X);
samer@4: