Some of the PAC functions require more / less timepoints in order to achieve a stable estimate. Perhaps it's worth adding a check for any given PAC function to make sure that users are doing something sensible with their estimation. E.g., "given the size of input data and the sampling rate of the data, raise a warning if we are below some threshold for n_timepoints". Thoughts?
Some of the PAC functions require more / less timepoints in order to achieve a stable estimate. Perhaps it's worth adding a check for any given PAC function to make sure that users are doing something sensible with their estimation. E.g., "given the size of input data and the sampling rate of the data, raise a warning if we are below some threshold for n_timepoints". Thoughts?