change distribution for random processes with lower limits

[michaelgaunt404]

most random processes use normal gaussian model to draw from.

This is not good because draws can surpass meaningful thresholds: boarding times less than zero.

Current solution is poor work around as it augments final distribution due to resampling threshold violating draws.

` library(tidyverse) samp = rnorm(1e5, 4, 4)

data.frame(raw = head(samp, 50000), aug = head(samp[samp>1], 50000)) %>%
pivot_longer(cols = c(aug, raw)) %>%
ggplot() + geom_histogram(aes(value)) + facet_grid(rows = vars(name)) `

gamma or weibull

[michaelgaunt404]

https://www.sciencedirect.com/science/article/pii/S2590198219300272 should look up literature on time it takes to board a bus