Allow peak estimation across groups

[zkamvar]

Given that find_peak() can do so without pooling and bootstrapping considers groups, it's not much of a stretch to think that we can bootstrap across groups and find the peak. The only issue is that the current implementation would have to be changed in a non-trivial way to accommodate. We would need to pre-allocate a data frame (not a matrix) of dates (if the input is date), and then we can run a for loop to add the computed bootstraps in with df[i, ] <- as.list(find_peak(boostrap(x), pool = TRUE))