*mutter mutter grumble*
I do a fair amount of work with epidemiological models of infectious disease. Both of the empirical regression type and also theoretical simulation models of disease transmission. The latter is the subject of this rant – the former may be the subject of later ones.
One type of model is a ‘compartmental’ model, usually called an SIR model (or some variation thereof). Model subjects are classified as being in a particular state (compartment) – usually most basically Susceptible (S), Infectious (I) or Recovered (R) – hence the name SIR, with formulas governing the flows between them. Normally, these models are written as a series of related differential equations, using some form of fairly complex software.
But for a presentation I’m working on, I wanted a very simple, very clear, illustrative model. Since I understand basic addition and subtraction more than I understand calculus, I decided to make a model using difference equations – equations in discrete rather than continuous time increments. Generally, for simple models, there’s practically no difference, and I find it much easier to much about with the formulas.
What I discover today though is that they’re a much bigger pain to put together. The nice, simple models have mind-breaking weirdness to some of the formulas, and doing them in a spreadsheet for an easy, pretty graph brings the graphing abilities of both Excel and Numbers to a halt. What could have been done in 10, 15 minutes at most using swanky software and calculus took the entire length of the vice-presidential debate. Because I did it the “easy” way.
Filed under: Epidemiology, Grad School Life | Leave a Comment