<< ica_blocs_signals (X) One matrix of observations/variables icajade >>

fact >> (X) One matrix of observations/variables > ica_dwresiduals

ica_dwresiduals

the number of independent components in ICA is determined from the Durbin-Watson criterium

Calling sequence

res=ica_dwresiduals(x,n_ics,(options));

Arguments

x:

a matrix (n x q) or a Div structure

n_ics:

maximum number of independent components

(options):

an optional structure containing the following fields:

options.center: 1 for centering, 0 for no centering (by default)

options.method: choice among: 'tall','wide','kernel','comdim' or 'normal' (by default)

options.partitions: an integer; for 'tall','wide' and 'comdim' methods only (by default = 1)

options.plot: 0 = no plotting (by defautl); 1 = lowest correlations; 2 = all

res.all_dw:

all the Durbin-Watson values

res.all_dw.d is a matrix of dimensions (n_ics x n)

res.mean_all_dw:

the means of the rows of res.all_dw.d

res.mean_all_dw.d is a vector of dimensions (n_ics x 1)

res.residus:

the residues; a list of n_ics Div structures

res.residus(i).d is a matrix of dimensions (n x q)

Examples

[my_result]=ica_dwresiduals(x,5,my_options);

Bibliography

Authors


Report an issue
<< ica_blocs_signals (X) One matrix of observations/variables icajade >>