Thu Jan 24 18:47:08 PST 2002This binary version of the runica() function of Makeig et al.  containedin the EEG/ICA Toolbox runs 12x faster than the Matlab version. It usesthe logistic infomax ICA algorithm of Bell and Sejnowski, with naturalgradient and extended ICA extensions.  It was programmed for unsupervisedusage by Scott Makeig at CNL, Salk Institute, La Jolla CA. Sigurd Enghofftranslated it into C++ code and compiled it for multiple platforms. J-RDuann has improved the PCA dimension-reduction and has compiled thelinux and free_bsd versions.To use the function, call with a ".sc" file argument. For individualCopy and modify the sample script "ica.sc".                % ica < myversion.sc Ica creates two files, "xxx.wts" and "xxx.sph" containing weights andsphere matrices such that                >> ICA_activations = wts * sph * data; The "xxx" stem in the output files may be specified within the input .scparameter file. See the sample .sc file for arguments, and the EEG/ICAtoolbox tutorial for more details.ica_bsd   - FreeBSD Unixica_linux - linux version (tested under Red Hat)ica_sgi   - SGI (older version, PCA not optimized)ica_sun   - Sun OS (older version, PCA not optimized)ica.exe   - PC version (older version, PCA not optimized, stable?)Scott MakeigJan. 26, 2002