lyngby_efir_main - Main exhautive FIR function
function [result_efir,numcoef] = lyngby_efir_main(PN, X, R,...
maxfilterlength, numreshuffle)
Input: PN Input (the paradigm)
X Target (the datamatrix)
R Run structure
maxfilterlength Filter length (lag)
numreshuffle Number of crossvalidation
permutations
Output: result_efir Filter coefficients
numcoef Optimal number of filter
coefficients
The FIR filter used in combination the SVD pseudo
inversion. The optimal model in each voxel is determined by
the Generalization error, based on the exhaustive search
through a 2D space. The 2D space is span by 'models' and
'splitratios'. The 'splitratio' is simply how the dataset
is split into a trainset and a testset. The 'models' is the
size of the SVD pseudo inversion length.
See also: lyngby_fir_main