POP_MULTIFIT Fit multiple component dipoles using DIPFIT

Usage: >> EEG = pop_multifit(EEG); % pop-up graphical interface
>> EEG = pop_multifit(EEG, comps, 'key', 'val', ...);

Inputs:
EEG   
input dataset.
comps   
component to fit

Optional inputs:
'dipoles'   
[1|2] use either 1 dipole or 2 dipoles contrain in symetry. Default is 1.
'dipplot'   
['on'|'off'] plot dipoles. Default is 'off'.
'plotopt'   
[cell array] dipplot() 'key', 'val' options. Default is 'normlen', 'on', 'image', 'fullmri'
'rmout'   
['on'|'off'] remove dipoles outside the head. Artifactual component often localize outside the head. Default is 'off'.
'threshold'   
[float] rejection threshold during component scan. Default is 40 (residual variance above 40%).

Outputs:
EEG   
output dataset with updated "EEG.dipfit" field

Note: residual variance is set to NaN if DIPFIT does not converge

Author: Arnaud Delorme, SCCN/INC/UCSD, La Jolla, Oct. 2003

See the matlab file pop_multifit.m (may require other functions)