Coding

As I explained my project to jwisha in an email today (and because I spent the entire day in the lab today), I thought it would be a good time to explain the basic idea behind my 10 week lab rotation. 

I am working on a model for citrus canker disease, which affects a variety of citrus species and significantly reduces yield. It can go through three stages – Susceptible (no disease but is susceptible to it), cryptic (has the disease and is capable of transmitting it, but shows no symptoms), and infected (has the disease and is capable of transmitting it, as well as showing symptoms). The infected trees do not die. The main issue is with the cryptic stage, as the disease cannot be detected at this stage and therefore the infected tree cannot be removed, but it can still transmit the disease to susceptibles. The trees are removed in their infected stage once the symptoms are detected. My model looks at two hypothetical species of citrus tree, one with low cost fruit produced but a fast rate of cryptic to infected, and one with a high cost fruit produced but slow rate of cryptic to infected. I am trying to find the trade-off between disease spread and profit gained.

I am slowly getting there, and it's actually really fun (when the code works)! These graphs don't actually show what my goal is but instead they are plotting the disease spread as an initial starting point. I've been working on more complicated coding over the past few days but the pictures for that are very boring at the moment so I thought I'd show these instead. 

For those who are interested:
Top LHS: Shows the spread of disease for susceptible, cryptic, and infected plants for two species (species 1 full line, species 2 dotted line) with two different infection rates. The purple lines are for removed individuals that were cut down to limit the spread of disease. 

Bottom LHS: the same as the top but with stochasticity added in (it makes the model more like "real life")

Both RHS: levels of infected individuals for two models repeated multiple types with a line going through the mean value of them - like the line of best fit. 

Phew. I hope that makes some sense!