My little idea -- using MATLAB to realize reading similar to ring buffer

Recently, a paper reappears , In the article, it is required to buffer the current time and before $N-1$ A measurement 、 And the corresponding filter variance matrix , I thought of the ring buffer , I wonder if my understanding is accurate , I created a length of $N$ The buffer , Then use an index number to correspond to the beginning of the buffer , But as we all know ,MATLAB All numerical data in are stored in a matrix , This makes me want to fit more dimensional matrices , This led to the following attempts , It's very easy to use ！ The main method is to use eval Function will :, The dimension of minus one stitching is converted into a command , So as to achieve any dimension $N$ Measurements only need to be input , You can output the corresponding values in the correct order .

function bufferValue = readBuffer(buffer, bufferStart)
    % buffer The last dimension of is index 
    colonStr = ':,';
    colonNum = ndims(buffer) - 1;
    colonJoinStr = repmat(colonStr,[1,colonNum]);
    bufferLastDimSize = size(buffer,ndims(buffer));
    bufferValue = zeros(size(buffer));
    for index = bufferStart:bufferStart+bufferLastDimSize-1
        indexTmp = index - (floor((index - 1) / bufferLastDimSize) * bufferLastDimSize);
        indexReal = index - bufferStart + 1;
        eval(strcat('bufferValue(',colonJoinStr,num2str(indexReal),')','=','buffer(',colonJoinStr,num2str(indexTmp),');'))
    end
end