Should we model growth as continuous or discrete

How to model without making nonsensical predictions.

UdeS
stan
Author

Andrew MacDonald

Published

November 11, 2022

library(targets)
library(ggplot2)
library(tidyverse)
library(tidybayes)

How do things grow (or: decay)

Let’s start with a simple simulation:

start_size <- 45
decay_rate <- .3
tibble(time = 0:20, 
       mass = start_size * exp(-decay_rate * time)) |> 
  ggplot(aes(x = time, y = mass)) + 
  geom_point()