kordinglab / nerds Goto Github PK

We should create setup_nerds.m file that include all path of nerds folder. I'm not sure whether we should include adjusted SPGL folder in NERDS since Eva adjusted and added few codes to SPGL.

We found the boundary effect of recovered spike train i.e. we tend to have excess spikes at the end of prediction. We are trying to see where in the code that give rise to this issue.

(1) We should add plot function for these complex outputs
(2) And handy function to convolve estimated spikes with learned template so user don't need to use convolve.

After solving inverse fft, we got circular adding of the signal. We can fix by padding. However, the have some wrong work flow that I apply after I pad the coming input signal. Moreover, output has excess 1 index more than input signal

I suggest that you move all functions from utilities and spgl1_extend directories to a new private directory. Matlab automatically discovers these functions without having to add the private directory to the matlab path, and as an added benefit, these functions are hidden from use except by nerds.