Clear Filters
Clear Filters

Issue running the RUNICA function in the EEGLab toolbox

22 views (last 30 days)
Hi,
I'm trying to use the function "runica" from the eeglab toolbox to decompose a fairly large data set. Now I've been testing it on a small sample of the data (5 channels) to figure out what do do but have hit a snag with the outputs.
From the command:
[weights,sphere,project,activations]=runica(datiatt([1:5],:));
I am getting the following outputs:
>Weights == 5x5 double, -6.485min, 8.232max
>Sphere == 5x5 double, -0.0225min, 0.0323max
>Project =[2.041472649833849e+04,1.057366806435951e+03,8.424862489971001e+02,5.5
08867487460723e+02,86.325649879637750] , 86.3256 min, 2.0415e+04 max
>Activations== [0.521181380438761;-
0.503483128865776;0.168416662178569;0.087787753657055;-
0.278149228696979], -0.5035 min, 0.5212 max.
The issues I'm having
  • Activations is not equal to 128*length of datiatt matrix.
  • When trying to compute the unmixing matrix so that I can view the components and re-run my psd analysis I am getting the error message:
W=weights*sphere "matrix dimensions do not match".
Without being able to create 'W' I can't go any further with the analysis.
Does anyone have any helpful hints or suggestions?
Thanks

Answers (1)

Walter Roberson
Walter Roberson on 26 Feb 2013
Please show the exact error message for the multiplication. Also, please show size(weights) and size(sphere)
Your list of outputs shows Weights and Sphere in capitals. Please cross-check upper/lower case use, especially for Sphere, as there is a function sphere that could possibly be invoked if Sphere (capital) had been used for the variable.

Categories

Find more on EEG/MEG/ECoG in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!