In `surv_cutpoint` the values of max_stat$stats is not the entire distribution of the original data

[jkylearmstrong]

I think it's very misleading, the values in max_stat$stats are not the distribution of the entire original dataset, it is probably just the subset of the values tested, there is perhaps some logic in maxstat::maxstat.test which specifies which values to test, it probably does not test every possible cut point, it might only test up to the lowest such that there are at least n in each group.

This is a very easy fix.

maxstat::maxstat.test$stats <- data[,var_i] # this will copy all rows of var_i into stats then the distribution and plots will align more with the original data.

[jkylearmstrong]

Fixed proposed in #678