Converting data contained in a cell array into a 4D matrix for input into a CNN for feature extraction

Hi All,

I am working with EMG data and trying to use a CNN to extract features from the raw signal

I currently have the latest version of Matlab 2021a

XTrain (453x1 cell array) each cell conatins a sample signal (8*213) so to clairfy that is 453 samples and each sample contains 8 channel sequence with 213 data points

From the reading i have done this needs to be converted in to a 4D matrix to be acceptable for a CNN

I thought his would work based on an example set of code i found:

XTrain = reshape(XTrain,[8,213,1,453])

but i am getting errors:

Number of elements must not change. Use [] as one of the size inputs to automatically calculate the appropriate size for that

dimension.

I am not really sure how to make this work now. I have tried using empty [] in place of some of the dimensions but this has not worked either.

Any help would be appreciated

Kind Regards

Answer 1

[KSSV]

XTrain = cat(3,XTrain{:}) ;
XTrain = reshape(XTrain,8,213,1,453) ;

[CHRISTOPHER MILLAR]

Hi KSSV,

Thanks for answering so quickly

I have tried what you have suggested but i get an error on the first part:

Error using cat
Dimensions of arrays being concatenated are not consistent.

This is my fault i forgot to mention that some of my samples are 8*173 in length

Will i have to make all these samples the same length before i can move forwad?

Kind Regards

[KSSV]

% Demo with dummy data 
XTrain = cell(453,1) ; 
for i = 1:453
    XTrain{i} = rand(8,213) ; 
end
XTrain = cat(3,XTrain{:}) ;
XTrain = reshape(XTrain,8,213,1,453) ;

[CHRISTOPHER MILLAR]

Brilliant

Thanks alot for your help

This worked perfectly for me now