Kilosort 2 fails on extremely large "spike" trains

[brandonStell]

KS2 throws an error when it finds a huge number of spikes:

        "Number of spikes before applying cutoff: 36449620",
        "----------------------------------------Error using gpuArray/eig",
        "EIG did not converge at index = 881672."

There is a very easy fix that doesn't appear in the docker and the docker is the only way to run KS2 on later versions of Ubuntu.

Would it be possible to add a container with a try and catch added to runTemplates.m in order to use the economy decomposition when huge spiketrains are found?

    try
        [A, B, C] = svdecon(UA);
    catch
        [A, B, C] = svd(UA,"econ");
    end

The economy-size decomposition removes extra rows or columns of zeros from the diagonal matrix of singular values, S, along with the columns in either U or V that multiply those zeros in the expression A = USV'. Removing these zeros and columns can improve execution time and reduce storage requirements without compromising the accuracy of the decomposition.
https://www.mathworks.com/help/matlab/ref/double.svd.html#d126e1780458

I added this issue to the KS2 repo but I don't have much hope of it being incorporated there now that they are focused on KS4+.

[brandonStell]

As expected there doesn't seem much hope for getting this small change incorporated directly into the KS repo:
MouseLand/Kilosort#945

So our only hope for continuing to use KS2 is to get this into one of your containers. Note that these lines will not change anything for other users unless they receive the error message above indicating that svdecon function failed, in which case it will fall back to the svd function:
https://www.mathworks.com/matlabcentral/fileexchange/47132-fast-svd-and-pca